D5 speed settings, Only 1 degree difference between high and low.

[momo56]

So I have 2 D5's in series with what I call a highly restrictive loop. Using intel burn test the cpu max temp after 20 runs seems to only increase by a degree when I lowered it from setting 5 to setting 1. Is this normal?

Pretty P.O'd considering I spent quite a bit on two d5 vario's and a dual pump top to to replace my old D5 b

 

[Metroid]

It depends, as you said you have 2 D5's which may increase the pressure to as much as 2x. If you had only 1 then the speed settings might have made the difference. The good thing about doing dual is the benefit of a silent system. It dumps +1 D5 watts into water, reason increased 1C, more pressure lower noise RPM.

Bottom line: There is nothing wrong.

 

[JeffyB]

A few things. 20 runs of ibt is not going to be enough to fully saturate a loop with heat and as far as i am aware in most situations flow rate has very little effects on cooling, so long as the water is moving. Gabe from swiftech tested the effects of flow rate on temps and found it to have very little effect, similar to the myth that bigger tubing will give lower temps. Having 2 pumps has the benefit of redundancy should fail. Incase you dont know alreay know always run both pumps at the same speed or you risk damaging one of them.

 

[Greebo]

Not strictly true JeffyB. I would be interested in seeing Swiftech's results as it is certainly different to Skinee's labs findings.

Although, I would agree with you to a point, sufficient flow rate is required and the recommended is 1.5g per minute. Once you go above that, the gains are reduced.

http://skinneelabs.com/ek-supreme_supremelt.html

He tests from one laing d5 on setting 1 (v low flow) to three MCP355 pumps on full speed (v high flow)

You can see on his cpu block reviews that there is 9 degrees difference between the very low (0.3g per min) to the very high flow (2.7 gal per minute).

Although also notice that between medium high flow (one MCP355 on full speed) and very high there is just under two degree difference.

So in the op's case, he probably did have sufficient flow with just one pump hence the one degree gain being shown.

In my case, with an loads of restrictive blocks and rads in one loop, I gained about 7 degrees as my flow rate with one pump was appauling and just was a trickle in the reservoir.

 

[shadowscotland]

I have to agree with Greebo
Above and below 'sufficent' flowrate there is a differance especially with impingement blocks.
but over 1.5gpm the improvement in temps is very limited if any.

momo what do you class as a restricitive loop?

 

[JeffyB]

Yes thats why i said in most situation. Compared to an average flow rate a higher flow will show relatively small gains, 0.3g is rediculously low and is well outside what i would call most situations. I personally see next to no difference in temperatures between speed 2 and 5 on my vario.

 

[TheCenturion]

I've got my vario set to setting 4. I'm going to turn it on for the 1st time later today. I know pumps generate a small amount of heat, but I was under the impression that it's negligible? (as you've demonstrated in the op)

 

[JeffyB]

I have skinnees pump and rad optimiser has values for heat dump, here you go

P1-2.5w
P2-4.35w
P3-6.51w
P4-9.89w
P5-17.34w

 

[shadowscotland]

Most passive or silent watercooling loop run at below 0.5gpm.
Where long loop length and 'quiet' pumps are the norm.

Most multi block / single pump loop run below 0.5gpm

But I accept many people over spec their pump for pump/block/rad/res/pump loops.

 

[Greebo]

Yes thats why i said in most situation. Compared to an average flow rate a higher flow will show relatively small gains, 0.3g is rediculously low and is well outside what i would call most situations. I personally see next to no difference in temperatures between speed 2 and 5 on my vario.

You say that but if you have a very restrictive loop, a d5 even on setting 5 can give a flow rate as low as that.

I did a rough measurement with mine and I was getting about 0.4g per minute on max setting. It was bad enough to make me think my pump was broken. I now have over 1.5g per minute and everything is fine again. It was so bad I struggled to pump air bubbles round the system and bleeding was a nightmare.

I had an swiftech xt block, two gtx470 waterblocks, two xspc rads and a very long loop. Total water volume in the loop was about 1.3 litres.

Now with two pumps, the water in the reservoir is like a whirlpool again on setting 5. In fact I am thinking of running both pumps on a lower setting again.

 

[TheCenturion]

I have skinnees pump and rad optimiser has values for heat dump, here you go

P1-2.5w
P2-4.35w
P3-6.51w
P4-9.89w
P5-17.34w

Thanks Paul, that's really helpful, especially as I'm going to have to keep a big eye on temps to begin with

 

[JeffyB]

You say that but if you have a very restrictive loop, a d5 even on setting 5 can give a flow rate as low as that.

I wouldn't recommend a d5 for that type of loop, you need more head pressure than a d5 can deliver.

 

[Greebo]

I wouldn't recommend a d5 for that type of loop, you need more head pressure than a d5 can deliver.

I realised that the hard way! All sorted and happy now with two d5's. Also swapped the xt for an extreme HF which helped as well.

 

[momo56]

Right today i tested 2 d5's (setting 5) for 2 hours under prime and then turned one of the d5's off for another 2 hours. Saw no change in temps :s

 

[Greebo]

Which must mean you had more than sufficient flow with just one????

 

[momo56]

hmm yhh just tried daul loops with the cpu with its own rad but used a DDC with res top. Only got a 2 degree improvement. Guess flow was enough with just a d5 :s

 

[momo56]

I have to agree with Greebo

momo what do you class as a restricitive loop?

sorry just saw this message now. My loop that I consider restrictive:

MIPS RAM BLOCK > mosfet 1 > mosfet 2 > EK supreme HF > NB/SB block > EK 4870x2 block > 3 Thermochill 120.4 rads > reservoir > 2 D5 pumps in series