Eheim 1250 + WW + Maze4 GPU = slow flow rate?

[GenGoku]

Hi,

I got the new core from dtek today (for those of you who read my dtek 'holy' core) and setup the loop again. after trying to bleed the system, i noticed that the water shooting out of the end of the hose doesn't seem to be going all that strong--kinda limp. is this normal because i could have sworn flowrate was a lot faster before.

*Warning 56kers*
Bleeding
Bleeding 2
Bleeding 3
Setup*
Sorry i couldn't Save-for-web it as photoshop is on my main pc (i'm stuck on my not-so-good laptop till the watercooling project is done)

ps. if anyone is wondering about my heatercore, i put cardboard infront and behind the core to protect it from any damage.

*This is an older "version" of how it is going to be setup. the changes are as follow

• Eheim outlet to top core barb
• T-cap will be in the case going up along the edge of the motherboard

[sn_85]

to help improve flow i suggest having the water flow straight through the T-Line to your white water instead of taking that 90 degree bend to your white water.

[GenGoku]

you mean have the cap of the T straight up? if so, that's how it will be. When i move the T-cap straight up, in the basin, the waterflow doesnt seem to be much stronger. i just edited my first post to show how it will look in the case.

please correct me if i'm wrong.

[edit]
i thought it could be air trapped in the core or eheim, but doubt it as i've already tried moving it around as much as i could--don't see any more air bubbles flowing out of it when moving it.

[edit2]
oh my gosh. i might have another doa core. i tried finding the bottle neck by testing the eheim by itself then with the eheim + core. i put the core outside of the basin so it's out in the air. i can shake water out the core.

[edit3]
yep. i think it is. this one is even worst than the last. i had to try a little hard to get some water to come out of the last core, but this one, i just shake it (by itself, filled with water, thumbs plugging barbs) and a whole lot of sprinkles come out.

i really hope i'm wrong and it's just water that accumulated while it was in the basin. i'll try again in the morning after the exterior has had time to dry out (leaving water inside) and do my "shake test" again.

[Cathar]

Do the T as suggested.

You should expect >6lpm flow rates with the Eheim 1250, the White Water and the heater-core. If not, then it will be the heater-core that is holding you back somewhat.

If you're achieving 6lpm (~1.6gpm) then it's all good. Very little to be gained by pushing flow rates higher than that with the White Water.

Oh, and don't submerse the core - they're not particularly designed to be submersed and it's totally unnecessary, and then you won't have to worry about whether when you shake it if it's water from inside or water from outside.

[GenGoku]

Originally posted by Cathar
Do the T as suggested.
<snip>
Oh, and don't submerse the core - they're not particularly designed to be submersed and it's totally unnecessary, and then you won't have to worry about whether when you shake it if it's water from inside or water from outside.

do any of my pictures show the T as suggested? if not, could someone please explain more on how it should be setup?

the only reason the core was submersed was because the outlet tube was relativly short (in a lian li pc65). it was easier to have the core inside the basin than leave it outside because of the way the parts are setup (see pictures above).

btw, i checked the core again (left outside for several hours w/ water inside--barbs facing up) and when i lifted up the core, there was a small puddle under the core.

[sn_85]

the flow should go straight not at 90 degrees. see how a T line has 2 ends that are opposites to each other and then there is an adjacent one that goes into the two. the single one that is perpendicular to the two other opposite ends. the single one should be where the airtrap/bleeding tube is.

[bigben2k]

No, the tee was setup ok. Setting it up so that the coolant runs straight through it would require a different placement. Either way, the difference is not significant, at this stage.


Boy, you're really not having any luck with those cores!

To test it, I use a tee, fitted with an air valve. You can simply run tubes to both connections of the core, and "inflate" it. DO NOT EXCEED 25 psi! You can run the test with the core full of water, or not. You can also use food coloring, in the water.

[GenGoku]

sorry all, i think i may have over-reacted. i checked it again this morning and the core was still full of water and it passed the "shake test." i'll try running the loop again but with everything out of the basin.

[bigben2k]

Hummm... I suspected as much. The fin shape can easily retain some water.

Still, you should be able to get better flow than what can be seen in the first pic (first post). Then again, that flow rate might be correct, it's really hard to judge without a measurement.

Since we know the Eheim 1250's curve, it's possible to determine the actual flow rate, with a simple (and cheap) manometer set up. Is this something that you'd be interested in doing?

[GenGoku]

all i need are directions (read: cheap)

btw, i tried making the T so the straight part goes to the cpu block but now the rest of that line has a weird shape. guess i'll have to cut more tubing to get it right. bigben2k, were you saying there's very little difference if i were to set it up this way?

thanks for the help all, i really appreciate it

[bigben2k]

At this stage, yes, it makes very little difference.

Otherwise, people have switched an Eheim 1048 for a 1250, and haven't noticed any improvement, so the whole tee thing, IMO, is really futile. But... also IMO, the tee needs to have the flow run through it at 90 degrees, to be able to extract some air.

On the other hand, running the coolant straight through should work, but the straight run would have to be horizontal, and as long as possible, to allow air to escape vertically, up the tee.

As for the manometer... set this up around your pump (with a complete loop, including core and block), and I'll post a curve of the Eheim pump on the next post.

[bigben2k]

Here are the Eheim curves (below).

Once setup, all you have to do is turn it on, and measure the height difference in the two columns of water. BTW, if your tubing run is too short, you'll have water on the floor.

Tke that measurement, and find it on the graph. draw a perpendicular line, from where your reading hits the 1250's curve, and you'll know approximately, what your actual flow rate is. (unfortunately, everything is metric, so a little math might be in order See www.convert-me.com )

[GenGoku]

could you give more details on the nanometer. i'm not too sure what it does

[GenGoku]

i dont think the core is the problem since making a loop (eheim->core->basin) the flow rate is very high.

[edit]
the problem is the WW. i think it's probably dirty

[edit2]
yep. just as i thought. there was a lot of gunk from the cardboard i used to protect the core's fins. will the discoloration have any impact on the WW's performance? (the pic makes it to look more gross than it really is)

[bigben2k]

There you go.


The manometer is rather simple. you hook up tees at the pumps inlet and outlet, so that you can run two tubes up vertically. Fill the whole system with water, until you get a level that sits in the middle of those vertical tubes. Then when you turn on the pump, one of the levels will drop, and the other will rise. Measure the difference in height, of those two levels: the result is the amount of pressure that your pump is supplying.

Keep in mind that the pressure could be as high as [1.5 meters water column], so the level should rise up and down accordingly. In other words, use a pretty long run of tubing, ~ 2 meters should do.

The tubing diameter is irrelevant.

Then lookup that height on the pump's curve (posted above) to see what the actual flow rate is.

[FuzzyFace]

Bigben2k:
I thought that a manometer was a single tube. Measuring the height of the water in the tube gives you the pressure in inches of water, or mm’s of water, or meters of water (whatever your favorite units of measurements are) in the system at that point.

Placing 2 manometers, one at the pump inlet and one at the pump outlet and measuring the difference in height gives you the pressure drop due to the resistance in the water block/s and heater core.

Am I all screwed up in my thinking or what?

[bigben2k]

A single tube would give you a reading relative to atmosphere pressure. You'd measure it against the level, when the pump is off. You'd still need a second tube, for the level to change.

In this case, we want to measure the pressure differential, between the pump inlet, and the pump outlet. So we set up a tube at both ends of the pump, and measure the height difference.

[redleader]

I have a low flow Maze2, a 1250 and i used a DTek core for a long tme. it should be flowing a LOT more then that. The 1250 is no beast but it handles loads a lot better then that.

[GenGoku]

Originally posted by redleader
I have a low flow Maze2, a 1250 and i used a DTek core for a long tme. it should be flowing a LOT more then that. The 1250 is no beast but it handles loads a lot better then that.

after taking out the gunk (pic above), the flow shoots out to the lower worm clamp of bleed2 (pic above). does that seem about right?

[bigben2k]

Yep, that's more like it.

(btw, it's Bleed1, not 2, right?)

[FuzzyFace]

Thank’s bigben2k. I know what a manometer is and what it does, and I understand the principles here. I just did not apply my knowledge properly. Although I have become interested in water cooling my computer (I am slowly working my way there) my 70 year old eyesight is giving out on me. So, I have to limit my reading of lengthy technical documents.

[GenGoku]

(new topic instead)

[bigben2k]

Originally posted by FuzzyFace
Thank’s bigben2k. I know what a manometer is and what it does, and I understand the principles here. I just did not apply my knowledge properly. Although I have become interested in water cooling my computer (I am slowly working my way there) my 70 year old eyesight is giving out on me. So, I have to limit my reading of lengthy technical documents.

No problem. Feel free to PM me, any time.