How much pump is enough? How much is too much?

[friku]

It´s the S-240 one. If this one could be undervolted should be nice,...

Cheers,
friku.

[nikhsub1]

Originally Posted by friku

It´s the S-240 one. If this one could be undervolted should be nice,...

Cheers,
friku.

That is the exact one I have, you should be able to get it down to atleast 15v or so. But, I would not run the pump at less than 18v, it is nearly silent there anyway.

[nightic]

Originally Posted by nikhsub1

That is the exact one I have, you should be able to get it down to atleast 15v or so. But, I would not run the pump at less than 18v, it is nearly silent there anyway.

How silent would you say the 18v "nearly silent" is, specifically in comparison to an MCP600/Aquaextreme @12V which is too noticeable for my taste?

[I am unfamiliar with the sound levels emitted by the D4 @5v & Eheim 1250 as described previously by Cathar.]

[nightic]

PS - In case anyone in the UK is considering buying one new from official sources - price quoted from both Iwaki UK and a distributor range from GBP£247-303 + carriage!

[friku]

Thanks Nikhsub1, can´t wait for having it installed... Probably in January I would have the system completely mounted...

Cheers,
friku.

[Cathar]

Originally Posted by nightic

How silent would you say the 18v "nearly silent" is, specifically in comparison to an MCP600/Aquaextreme @12V which is too noticeable for my taste?

[I am unfamiliar with the sound levels emitted by the D4 @5v & Eheim 1250 as described previously by Cathar.]

Much, much quieter than an MCP600/50Z @ 12v. MCP600/50Z is quite noisy by the metrics of the silence focused crowd.

[Cathar]

Did a heat-dump test on the RD-30. It's heat dump into the loop is actually fairly high, despite being a magnetic armature design. This is supported in that the casing barely gets warm.

Basically measured that around 83% of the pump's power draw gets passed into the water during in-line use.

Pump was measured to draw 28.1W. Calculated heat dump was 23.2W, with around an estimated 10% margin of error on that value.

Plugging that back in to the maths of the first two posts, this works out to:

+20.7C => Iwaki RD-30 @ 18.0v
+20.8C => 2 x Swiftech MCP600 @ 13.8v (series)
+20.9C => Swiftech MCP600 @ 13.8v
+21.1C => Iwaki MD-20RZ @ 60Hz
+21.1C => 2 x Swiftech MCP600 @ 12.0v (series)
+21.2C => Iwaki MD-20RZ @ 50Hz
+21.4C => Swiftech MCP600 @ 12.0v
+21.4C => 2 x Eheim 1048 (series)
+21.5C => Iwaki MD-15R @ 60Hz
+21.6C => Swiftech MCP650 @ 12.0v
+21.6C => Iwaki MD-30RZ @ 50Hz
+21.7C => Iwaki MD-30RZ @ 60Hz
+21.9C => Eheim 1250
+22.3C => Eheim 1048
+23.3C => Eheim 1046

[nightic]

Now that a little more is known of the DDC, any chance you could place that in the list also?

[Cathar]

Originally Posted by nightic

Now that a little more is know of the DDC, any chance you could place that in the list also?

Using an estimated 8W of heat dump for the Laing DDC.

+20.7C => Iwaki RD-30 @ 18.0v
+20.8C => 2 x Swiftech MCP600 @ 13.8v (series)
+20.9C => Swiftech MCP600 @ 13.8v
+21.1C => Iwaki MD-20RZ @ 60Hz
+21.1C => 2 x Swiftech MCP600 @ 12.0v (series)
+21.2C => Iwaki MD-20RZ @ 50Hz
+21.4C => Swiftech MCP600 @ 12.0v
+21.4C => 2 x Eheim 1048 (series)
+21.5C => Iwaki MD-15R @ 60Hz
+21.6C => Swiftech MCP650 @ 12.0v
+21.6C => Iwaki MD-30RZ @ 50Hz
+21.7C => Iwaki MD-30RZ @ 60Hz
+21.9C => Eheim 1250
+22.1C => Laing DDC @ 12.0v
+22.3C => Eheim 1048
+23.3C => Eheim 1046

[nightic]

Going to be cheeky/lazy and ask where the oft-quoted (by Laing themselves) 13.2V (1.75GPH/16' head) figure would reside.

[Cathar]

Originally Posted by nightic

Going to be cheeky/lazy and ask where the oft-quoted (by Laing themselves) 13.2V (1.75GPH/16' head) figure would reside.

Using an estimated 11W of heat dump for the Laing DDC @ 13.2v

+20.7C => Iwaki RD-30 @ 18.0v
+20.8C => 2 x Swiftech MCP600 @ 13.8v (series)
+20.9C => Swiftech MCP600 @ 13.8v
+21.1C => Iwaki MD-20RZ @ 60Hz
+21.1C => 2 x Swiftech MCP600 @ 12.0v (series)
+21.2C => Iwaki MD-20RZ @ 50Hz
+21.4C => Swiftech MCP600 @ 12.0v
+21.4C => 2 x Eheim 1048 (series)
+21.5C => Iwaki MD-15R @ 60Hz
+21.6C => Laing DDC @ 13.2v
+21.6C => Swiftech MCP650 @ 12.0v
+21.6C => Iwaki MD-30RZ @ 50Hz
+21.7C => Iwaki MD-30RZ @ 60Hz
+21.9C => Eheim 1250
+22.1C => Laing DDC @ 12.0v
+22.3C => Eheim 1048
+23.3C => Eheim 1046

[nightic]

Thanks Stew, extremely useful.

I do however have a bit of a dilemma which I'm sure some in my position share;

Whilst chasing performance I'm also intolerent of particularly forward noises.
For this reason I've been unable to get to grips with the Aquaextreme/MCP600 even though I've taken steps to ensure that it's especially well isolated.
I grant you it has not been run in for the 24hours+ after which noise levels are reported to drop as my observations in the first few hours of use indicate that it's never going to be quiet enough for me personally.

From your comments and figures it would seem that the RD-30 is close to ideal apart from the requirement of an ancillary PSU/DC-DC convertor.
Of course these pumps are not proving to be easily or regularly sourceable at a reasonable expense.

If one is to ignore the Eheims for reasons of compromised 'modern' performance (mediocre head, for what it's worth) and convenience (mains powered/potential EM interference issues/size [1250]) then the DDC/MCP350 begins to look increasingly attractive.

Now I don't wish to antagonise BillA beyond his default level (j/k) but there does seem to be a market for 2x these pumps run in parallel.

Assuming, going on the scant reviews thus far published, that the pump @12V is close to inaudible (if not entirely so) then this combination should provide at least D4/MCP650 (12V) levels of performance with far more comfortable noise levels.
The cost aspect for the discerning enthusiast is perhaps not as critical as one would think, given that many are happy to incur extra expense & effort to overvolt their 12V pumps for marginal performance increases.

My personal concerns (not suggesting that they are valid issues, they are merely concerns) with running the single pump in an already plumbed-in 1/2"ID loop with a higher-than-average length of tubing (~10' -- external radiator) are that despite the high head, the relatively low flow may unnecessarily hinder performance.
A switch down to 3/8"-7/16" ID tube may be preferable in this instance but involves considerable effort whereas plumbing in 2x DDCs would merely require a few reducers, wyes & short length of 3/8" ID tube.
There is also the psychological hurdle of contemplating a 6.5mm ID pump amongst a mass of 12.5mm ID PVC.

Just how wrong is it to covet that 2nd DDC?

[Cathar]

A second DDC may well be a good idea, and given its characteristics it would generally be desirable to put them in parallel as you suggest. In-series would suit the incredibly high flow-resistance blocks (i.e. even the NexXxoS XP will only be borderling restrictive enough from a cursory examination of the curves to warrant an in-series setup).

In parallel at 12v, we are getting a curve that looks much like the MCP600 @ 13.8v, with heat dump being a little higher, but also slightly higher head. Haven't worked it out, but from a cursory glance it looks like using the MCP600@13.8v figures would be a very close estimate.

In parallel @ 13.2v, which is very easy to achieve even with a small DC-DC converter which can be bought for ~$30US or so, and the putting them in parallel looks something like the 2 x MCP600 @ 12v in series curve, but with more heat.

Generally speaking depending on how you set it up, you'll be hanging around the 21.0-21.3C mark sort of positioning in the above estimates.

I don't think it's wrong to covet 2 x DDC's, but like the scenario's describe, don't go expecting big gains. Don't let a few converters bother you. Generally better though to put some 3/8" ID (1/2" OD) tubing on the pump barbs, and then slip 1/2" ID tubing on top and clamp it on. Removes the need for flow restricting reducers and fairly handily converts the pump into 1/2" OD barbs. Only worry if there's enough room to do so. May not be possible with fat 3/4" OD tubing, but 5/8" OD tubing (7/16" ID) may work.

I had originally started this thread to figure out at what point pumps became "too big", but it sort of evolved into a quest for the best balanced single pump. I understand the enthusiast point of view of chasing even the last 0.5C, but I also want it in a single compact pump, and the DDC is not that, although if using 2 pumps it does come somewhat close-ish, at the extra hassle of routing and wyes as you say.

[Chew_Toy]

What would the DDC be like in the 15~16v range?

This was taken from DDs web site.

– Yes. The pump will operate at a minimum of 6 volts and can operate at a maximum of 16 volts in a continuous duty environment. As voltage levels increase beyond the nominal 12 volts, life expectancy will decrease.

I would guess that the DDC 2@16V would be as or more reliable than the MCP600?AQX 50Z would at 13.8

[nightic]

Much obliged Stew, that was exactly what I was after.

One is due with me early next week, will be interesting to see how it does, moreso with the 25W 11-16V Meanwell DC-DC convertor bought (on your learned recommendation ) in preparation for use with the now superfluous Aquaextreme.

Only question that still remains is whether there's any advantage to be gained by switching the tubing over to 3/8" ID for a greater sense of continuity/lower total mass of water?

[Cathar]

Originally Posted by Chew_Toy

What would the DDC be like in the 15~16v range?

This was taken from DDs web site.


I would guess that the DDC 2@16V would be as or more reliable than the MCP600?AQX 50Z would at 13.8

I would estimate that 2 in parallel at 15.0v would roughly approximate the 60Hz MD-20RZ curve, and also with a similar amount of heat dump.

Munching it through a bit and it seems like the DDC really seems to plataeu with respect to heat dump vs flow rate increases with voltages above 12.0v. i.e. there doesn't seem to be much point in over-volting it.

2 in parallel at 16.0v would roughly approximate the 60Hz MD-30RZ curve as an estimate. Again with an estimared fairly high heat dump approaching 35W or so for the two pump. Again we'd be seeing overall performance plataeu, or even degrade due to the extra pump heat.

Seems like no matter what, the estimates are showing that the two pumps in parallel are plateauing out at around +21.0 to +21.3C, regardless of the over-volting performed.

Really it looks like one would have to go to a stronger radiator setup to make over-volting the DDC's above 12v worth-while.

[The Doors]

Just courios to know how OASE pumps (Alphacool branded) performs, especially the AP1500, any info about?

It seems to use 8W at 36dBA as reported on the AlphaCool website.

Thx,
Adriano

[Sunwong]

Arrrrg... I used a multimeter to test if my Meanwell S240-24 was correctly setup at 24V but did not test the lower and upper limits... I guess I will have to get the Power supply to work again to verify! :shrug:

I have not had my RD30 (Thk. CoolROD ) installed as of yet, but I was initially going for an undervolted setup so I will post figures whenever I get it done (believe me, that external case I am building is killing me )

I am pleased however to see that P-Q curve at 18V is so respectable, I had some doubts about the performance of the pump when undervolted. Plus, heat dissipation is incredibly low...!!!

Guess we did some nice shopping

Regards,

Sunwong

[Cathar]

Originally Posted by The Doors

Just courios to know how OASE pumps (Alphacool branded) performs, especially the AP1500, any info about?

It seems to use 8W at 36dBA as reported on the AlphaCool website.

Thx,
Adriano

Says 17W at Alphacool's website here.

[Les]

Originally Posted by Cathar

Plugging that back in to the maths of the first two posts, this works out to:
.......
+23.3C => Eheim 1046

But it is a much flawed "munching" machine
Phaestus's data* should be used to estimate "C/W"cpu and not an estimate of "C/W" for an insulated 10x10mm die.
A correction of 1/2QCp should be applied to Bill's radiator data.(0.5x1/QCp - data is LMT wet-side)
The correct W(pump) for Bill's radiator data(after!/2CpQ correction) is W(pump) at dead head(zero flow)

These details matter when calculating to 0.1c.

* Even using this is problematic,the Wrad should be corrected for the enigmatic variation(with flow rate) of W from CPU to WB.

[Uncle`BuZZ]

Originally Posted by Cathar

Using an estimated 8W of heat dump for the Laing DDC.

+20.7C => Iwaki RD-30 @ 18.0v
+20.8C => 2 x Swiftech MCP600 @ 13.8v (series)
+20.9C => Swiftech MCP600 @ 13.8v
+21.1C => Iwaki MD-20RZ @ 60Hz
+21.1C => 2 x Swiftech MCP600 @ 12.0v (series)
+21.2C => Iwaki MD-20RZ @ 50Hz
+21.4C => Swiftech MCP600 @ 12.0v
+21.4C => 2 x Eheim 1048 (series)
+21.5C => Iwaki MD-15R @ 60Hz
+21.6C => Swiftech MCP650 @ 12.0v
+21.6C => Iwaki MD-30RZ @ 50Hz
+21.7C => Iwaki MD-30RZ @ 60Hz
+21.9C => Eheim 1250
+22.1C => Laing DDC @ 12.0v
+22.3C => Eheim 1048
+23.3C => Eheim 1046

According to this graph :

the ranking depends of the waterblock in loop... In particular, if you watch the results of laing DDC, you can see his rank is different for the MCW6000 where it is worse than EHEIM 1250 (as your ranking) but for the nexxos XP, it's best than EHEIM1250 and Laing D4 (MCP650), even if figures are for 8w instead of 11w.

[Cathar]

Originally Posted by Uncle`BuZZ

the ranking depends of the waterblock in loop... In particular, if you watch the results of laing DDC, you can see his rank is different for the MCW6000 where it is worse than EHEIM 1250 (as your ranking) but for the nexxos XP, it's best than EHEIM1250 and Laing D4 (MCP650), even if figures are for 8w instead of 11w.

Indeed, ranking is highly dependent upon a number of factors:

1) Block in the loop (flat performance curve blocks, and/or highly restrictive blocks favor higher head pumps, while low resistance blocks favor higher flow pumps)
2) Radiator heat dissipation capacity (highly performing radiator will favor stronger pumps, weak radiators will favor lower heat dump pumps)
3) CPU heat load (lower heat loads favor smaller pumps)
4) CPU die size (affects what the block does)
5) Optionally GPU heat load - a hot secondary GPU with waterblock will tend to tilt favor towards stronger pumps too as the total CPU + GPU heat-load is more significant than the pump heat.

The rankings presenting in this thread are purely the result of a theoretical exercise. I chose the Cascade to work around mostly because it's a bit of a middle-ground sort of block. It's neither high restriction nor low restriction, and it possesses neither an extraordinarily flat C/W curve, nor an extraordinarily steep curve.

By all means, do not take these results as gospel. The results are highly setup dependent. This is more meant to be a rough guide to:

1) Seperate out the clearly weak pumps (Eheim 1046 for example) from the apparantly exceptionally balanced pumps.
2) Hopefully cause people to take pause before clicking "buy" on some 100W pump thinking that it'll improve their temps.

There are plenty of limitations to what's being presented above. Nobody should be thinking this is a one-size-fits-all sort of thing.

[JWFokker]

So what you're saying is that the pump should be matched to the system? So an MD30 might be more applicable to a dual CPU/SLI setup heavily overclocked than a single CPU loop at stock speeds? Awesome. I guess though, block design does play a bit of a role in there too. That Nexxos block does awfully well with an MD20RZ. Any chance you might include a Storm block in there at any point, being that you have one (or more) on hand?

[The Doors]

Originally Posted by Cathar

Says 17W at Alphacool's website here.

Absolutly right Cathar, but take a look to English version page here

In any way, what ya think about this pump?

[Cathar]

Originally Posted by The Doors

Absolutly right Cathar, but take a look to English version page here

In any way, what ya think about this pump?

The parent english page to that, which is here, lists 8W power consumption for the AP700, AP900 and AP1500. Clearly in error.

I have no experience with the pump. The various models are all just the same pump, but being run at different voltages. The AP1500 is just an over-volted (17V) version of the 12v AP900, and from what little I know about it, is actually fairly noisy at 17v, but that's just second-hand knowledge.

If we assume 3m of peak head, peak flow of 25LPM and 17W of power consumption, this pretty much places the pump on a near equal performance footing with the muffled Laing D4's.