Does a single EK D5 PWM pump have enough power to push through this setup?

abui1806

abui1806

abui1806

Associate
Joined
5 Mar 2014
Posts
1,129
Location
Southampton
Hi guys,

I am planing to watercool my system but I was wondering about this problem.
My system will have the followings:
1x CPU Block (EK supremacy EVO)
2x GPU Block
1x 120mm Radiator
1x 360mm Radiator
1x 480mm Radiator
1x Flow indicator

Thank you all for spending time in here. :)
 
If you want to hit 1 gallon per minute on your flow rate the no it won't be enough.

You'll be around 0.4-0.5 gallons per minute on your flow rate.

You won't loose much performance but bleeding and getting the air out will be a complete pain as the pump won't have a high enough flow rate to force the air out of the system.

An EK DDC with PWM will be a better choice and will give more performance.

Ditch that 120mm radiator to as they're pointless in loops with big radiators. This will save you a bit money and reduce your loops restriction. There will be no measure able performance drop by ditching it either.
 
almighty15 said:
If you want to hit 1 gallon per minute on your flow rate the no it won't be enough.

You'll be around 0.4-0.5 gallons per minute on your flow rate.

You won't loose much performance but bleeding and getting the air out will be a complete pain as the pump won't have a high enough flow rate to force the air out of the system.

An EK DDC with PWM will be a better choice and will give more performance.

Ditch that 120mm radiator to as they're pointless in loops with big radiators. This will save you a bit money and reduce your loops restriction. There will be no measure able performance drop by ditching it either.
Click to expand...

Thanks a lot for you help man. I will spend a good amount of time to think about this point.
 
I have more or the less the same set-up.

Only I have a 2x 360mm radiators and also water cool a cpu and 2x GPU's.

I currently have an EK D5 with PWM and that maxes out my loop at 0.6 gallons per minute.

I have an EK-DDC with PWM but haven't installed it yet but I know that it's be a good 50%+ more powerful then the D5.
 
1 gallon per minute is 273 litres/hour. Is that not massively overkill?!
I'm running an 18W DDC pushing through 8 Alphacool Cape Cora passive rads (in parallel), an EK 3x120mm rad, an EK Supreme HF CPU block, a Heatkiller GPU block and a flow meter. At full power it does 150l/h through this loop which is only just over 50% of that target. My temps are perfectly fine for gaming when running it at 7V which drops the flow to about 75l/h which is abour 0.27g/m.

When bleeding, my biggest problem is that the rad is top-mounted with the fittings on the underside so air gets trapped above them. Faster flow makes it easier to dislodge this with (heavy-handed) rocking but it also means that more of the air bubbles pass through the res to go round again. A slower flow (but not too slow) means that more of the air bubbles out in the res.
 
@almighty15: Hidden is good...it's when it leaks all over the carpet that it's a problem! ;)

I'm mainly after silence if I'm honest. It's a lash-back after a Shuttle small form factor PC with constantly variable insanity-causing device fan.

GPU is a 580GTX and there's not a massive amount of point in overclocking it further for literally a couple of fps.

CPU is a 2700K and I've overclocked the turbo to 4.2GHz so it boosts when I demand the power but runs cool when I don't. I know it'll happily run at 4.2GHz all day as I followed an SSD optimisation guide fairly recently and didn't realise that a side effect of changing Windows' power profile to 'High Performance' was that it pegged the CPU at 100%. Was thinking the bios had somehow changed settings until it suddenly dawned on me :rolleyes:
Was thinking of changing it recently but there's no massive advantage in changing at the moment so I'll probably wait until about 2016 when the interesting new stuff should be coming out :D
 
I run a D5 through this:

Cpu block
mosfet block
2 gpu block
tube res
2x 480 rad
2x 240 rads

havent had any problems. cpu clocked at 4.4 and 2 gpus overclocked. just run the pump at full speed for a couple of days to get the air bubbles out. i now run my pump at half speed.
 
yeah a d5 will easy run that. Anybody that says it won't has never tried it. Sure it might not have the initial umpfh to nuke the bubbles out in one swift motion but once you have it bled it will be fine (pretty much the same setup but more rads i found the difference between a quick bleed and not was the 2nd gpu block, when i went crossfire i had to spend a bit more time bubble slaying)
 
Last edited:
I run a dual D5 for that loop, down to a few reasons.

First, having fluid moving in the loop and having fluid flowing in the loop are two different stories. A radiator without a sufficient flow rate will not perform at it's best, period. If you don't have the sufficient pump power to maximize each rad, then why don't you decrease the rad area and save some money? Increasing rad size blindly is non-sense.

2. Redundancy is always better. If one D5 fail for any reason, it's always better to have some flow in the loop before you notice the fault and replace the pump.

3. Taken from 1, if you don't have enough flow rate, then you need to ramp up fan speed to compensate for cooling capacity. And I believe if you even increase 200rpm in each fan on all those rads, that will be a lot louder than having another D5 in the loop.



Bottom line, don't blindly go for more rads. Watercooling is not only about rad size, flow rate is also important.
 
Lol.

You will never get 50% more power over a D5 with a DDC.

I had 2 24v D5's powering thru 3x480s,2 CPU blocks,3 GPU blocks and 2 Mobo blocks,I took one out for servicing and had the other at 12v,the temp difference at the CPU was....1.3c.

Flow rates are massively overrated in regards to performance,the target you are looking for is 1 GPM.
 
I have 3x D5's, 3x DDC's and a Koolance PMP 500.

This is all connected to a Aquacomputer 5 XT with a flow meter and in my loop a 18w DDC was hitting 50% higher flow rates then a D5 on full speed.

If I dropped the flow rates my temps rise by an easy 10c.

Everything in my loop is measured with physical sensors and not in-accurate software sensors so I the actual temperature of my parts and water and every stage of my loop.
 
almighty15 said:
I have 3x D5's, 3x DDC's and a Koolance PMP 500.

This is all connected to a Aquacomputer 5 XT with a flow meter and in my loop a 18w DDC was hitting 50% higher flow rates then a D5 on full speed.

If I dropped the flow rates my temps rise by an easy 10c.

Everything in my loop is measured with physical sensors and not in-accurate software sensors so I the actual temperature of my parts and water and every stage of my loop.
Click to expand...

Aquaero eh? And you calibrated the sensors right?

50% more flow does not equate to 50% more power either,you have a restrictive loop tho that benefits from the higher head of the DDC.

And what temps are you talking about? Water? CPU?

Also...

almighty15 said:
Got 2x 18w DDC's and 2x D5's and thinking of doubling them up as my loop is very restrictive.

I want over 1 Gallon per minute on my flow rate, a single DDC manages 0.9 Gallons per minute and a single D5 on full power only managers 0.6!!

Now as I have a pair of each pumps I'm trying to decide which 2 I should run in serial to get my flow rates up.

The DDC's have more pumping power but the D5's have a higher flow rate.

I can get the pump top for the DDC's for £30 so while the top is cheap the pumps are nosier then the D5's.

On the flip side the top for the D5's will be more expensive but they'll be quieter.

What would you guys go for?
Click to expand...

Thats 33.3%.....not 50% like you claim.

And to answer that question,2 D5 strongs and a 24v controller will kick those little DDC pumps into the weeds.
 
Last edited:
Ok guys, thanks a lot for your support and knowledge. Admittedly, this is my first ever custom loop build so forgive me if I miss it somewhere. My target is not mainly a monstrous flow rate or a dreamy temp but I rather focus on the silent and aesthetic side of thing. The case I am planning to base upon is Caselabs SM8. So plenty of adjustments if needed.
For the CPU, I am aiming for 4.5Ghz on my i7 5820K only and some mild overclock on my SLI GTX 970 HOF.
 
abui1806 said:
Ok guys, thanks a lot for your support and knowledge. Admittedly, this is my first ever custom loop build so forgive me if I miss it somewhere. My target is not mainly a monstrous flow rate or a dreamy temp but I rather focus on the silent and aesthetic side of thing. The case I am planning to base upon is Caselabs SM8. So plenty of adjustments if needed.
For the CPU, I am aiming for 4.5Ghz on my i7 5820K only and some mild overclock on my SLI GTX 970 HOF.
Click to expand...


Then its a D5 hands down.

A PWM pump is handy but remember to step the duty curve,the pump hunting the rev range is massively more noticeable than a static pump at a higher speed.

Personally,im one to set to the highest speed tolerable soundwise and leave it at that.
 
You must log in or register to reply here.
Back
Top Bottom