laing d5 pump speed

JeffyB · 13 Nov 2011 at 13:19

Gabe from swiftech tested the effect of flow on temps a while back over at xs, basically so long as the water is moving it makes very little difference as there is such good flow in modern watercooling components.

Cleeecooo · 13 Nov 2011 at 13:20

JeffyB said:
Gabe from swiftech tested the effect of flow on temps a while back over at xs, basically so long as the water is moving it makes very little difference.

Is that still true if you have 4 D5's at 5?

JeffyB · 13 Nov 2011 at 13:43

Heres the pdf as the original thread has lost it's pictures.

http://0311a53.netsolstores.com/Res...ual influence of flow rate on system temp.pdf

As for 4 d5's on 5 who would run such a setup anyway? Thats also a lot of heat dump from the pumps. Dave explained it well.

1uke · 13 Nov 2011 at 14:54

Is this right the more Radiators/restrictive things in the line the higher the flow needed? Only asking as I was thinking of having 3 radiators 420,240 and a 140 in my next build.

Jokester · 13 Nov 2011 at 15:20

As long as you are over about 0.5gpm, anything more than that will only give you a degree or two at most.

RomeoS · 13 Nov 2011 at 15:22

I found it is entirely based how restrictive your loop is.

As an example:

I am running 2x D5's with the EK tops on the CPU loop.

CPU Block is a XSPC RayStorm and there are 2x 120.1 rads. Initially these were the Black Ice Stealth GTS120 but now are XSPC RX120.1's.

Running both pumps on "3" setting with IBT maxing out 8 threads on the CPU (now a 2700k running at 5Ghz). Changing the pumps from "3" to "5" instantly lowers the load CPU temp by 2C. Cofirmed with Open Hardware Monitor, CoreTemp and RealTemp.

Switch it back to "3, temps instantly go up 2C. Tried this several times with the same result.

This loop is obviously designed for high flow. I am hoping to get my hands on the new Swiftech GTZ block soon. These are MUCH more restrictive so the extra flow is going to be crucial to maintaining a decent temperature.

The myth that a slower flow rate is better as the water has longer to cool down has been around as long as watercooling and pc's themselves. Its all about finding the balance based on the radiators, blocks, pumps and fans you have.

EDIT To summarise, I run both pumps on CPU and GPU loops (4x D5's) all on "5" on foam paddings and they are inaudible.

JeffyB · 13 Nov 2011 at 18:41

2 degrees difference is nothing though and won't allow you to overclock any further.

Cleeecooo · 13 Nov 2011 at 19:04

Yeah^^
Onl worry if it's like 10 degrees difference

RomeoS · 13 Nov 2011 at 19:25

From my loop testing, every 125Mhz increase in overclock = 1C higher load temps. This is factoring in voltage bumps.

Your only likely to see 10C difference if you have a very restrictive loop.

My point is that adjust your pump speed based on your what you are cooling and your equipment. You will get to a point where the extra lp/h wont make much of a difference as other factors will be limiting you.

Jokesters advice of maintaining about 0.5gpm is spot on for the average WC loop.

Cleeecooo · 13 Nov 2011 at 21:09

What is gpm?

Jokester · 13 Nov 2011 at 21:12

Gallons per minute.

Cleeecooo · 13 Nov 2011 at 21:16

Jokester said:
Gallons per minute.

Ah, thankyou

So roughly 2.5 lpm?

Jokester · 13 Nov 2011 at 21:32

Cleeecooo said:
Ah, thankyou

So roughly 2.5 lpm?

Actually that'll be US gallons on Skinnee's, so it'll be less than 2lpm.

Cleeecooo · 13 Nov 2011 at 21:52

Damn yanks!!

shadowscotland · 13 Nov 2011 at 22:12

gallons per minute - doh! way to late with that reply

lanzarobbie · 14 Nov 2011 at 18:38

Whats the best way to to work out your gpm.

shadowscotland · 14 Nov 2011 at 18:56

an actual flow meter - or failing that 'MartinsWaterCoolingFlowRateEstimatorv3_1'

Jokester · 14 Nov 2011 at 19:13

http://martinsliquidlab.org/pump-and-radiator-optimizer-spreadsheet/

This is your best bet without a flowmeter.

joxang · 14 Nov 2011 at 19:31

No measurably difference to temps between 3, 4 and 5 for me. Two GPU blocks, one CPU, two rads. I run it on 5 as it's still silent, and my flow-meter looks nice when it spins quickly

.

debs3759 · 14 Nov 2011 at 19:44

shadowscotland said:
linky to my testing from a while back.

According to those figures, even at the maximum (5) setting, and assuming your figures are UK gallons, the D5 with a Laing top version 2 only pumps at 543 l/h. If your figures use US gallons it's even worse - 452 l/h!

The product specs at OCuK state 1200l/h as the maximum flow rate. Is it your loop that is restricting it that much, or is it possible that your flow meter was not giving accurate figures?

I'm asking as the loop I am building will have 2 D5 with top v 2 in series (for redundancy and extra pressure), and I am using a chiller that recommend a minimum of 1200 l/h flow rate. I need to work out if I should just buy a rad until I have the second PC ready to put in the loop.