Liquid cooling or not

[martin-w 3,235]

Deleted

[PigsInSpace 21]

 

Interesting mod to remove the rear fan and grill, and mount the rear case fan on the D14.

 

How much better would you say that is, compared to the standard set up?

 

 

It has actually been sized to accommodate another 150mm Nochtua.

 

Two benefits, roughly 33% larger area for air to move thru and without the added turbulence thru the mesh the actual gain is more, plus less static pressure=less noise.

[martin-w 3,235]

But precisely what?

 

What was the temp with the standard D14 set up and what was the temp after your mod?

 

Your centre mounted fan overlaps, so the greater fan size is only contributing to motherboard cooling, as the extra air flow isn't passing through the heat sink.

 

Same for the rear 150 fan, the overlap isn't cooling the CPU, the the overlap is only cooling the enclosure.

 

So for the reasons stated above, I don't think CPU temp would be that much lower than the standard D14 set up.

[PigsInSpace 21]

But precisely what?

 

What was the temp with the standard D14 set up and what was the temp after your mod?

 

Your centre mounted fan overlaps, so the greater fan size is only contributing to motherboard cooling, as the extra air flow isn't passing through the heat sink.

 

Same for the rear 150 fan, the overlap isn't cooling the CPU, the the overlap is only cooling the enclosure.

 

So for the reasons stated above, I don't think CPU temp would be that much lower than the standard D14 set up.

 

"Precisely" what test and for exactly how long bro? 1 hour? 24? what RPM, ambient tmp?  <_<

Every test run the temps were sometimes only marginally better and other times sometimes by more than 2-3 C, depending on the CPU draw and length of the test but >never lower than with just two, though I will say that two 140's plus the one 120 perform nearly as well in keeping the tmps down, yet they were a tad more noisier as expected.

 

No, the 150cm fan is not an oval, :blink:  it sits in precisely the same location as the 120 or 140 would sit so it has no where else to go but up, so it also has expanded cooling capacity to the sides as well as to keep moving cooler air thru, around the cooler and out of the case.

 

It works.

[martin-w 3,235]

Quote

"Precisely" what test and for exactly how long bro? 1 hour? 24? what RPM, ambient tmp?

Standard stuff when we test. Prior to mod run a stress test, full load, full RPM, determine average temps across the cores. Make a note of ambient temp. Keep all variables constant. Make a note of the delta T. Run the same test again after the mod. Time isn't critical as long as it's the same for each test, and provides sufficient time to reach max temp.

 

If we don't do this, and keep accurate notes, we have no idea.

 

 

Quote

sometimes only marginally better and other times sometimes by more than 2-3 C,

 

1-2 degrees is not significant statistically, as it's within the margin of error we would expect when testing any cooler. 2-3 degrees is minimal. So yes, pretty much what I expected, not much of an improvement in terms of temp compared to the D14 standard set up.

 

 

 

Quote

No, the 150cm fan is not an oval, :blink:

 

 

No, it's certainly not oval. :smile:

 

 

Quoteit also has expanded cooling capacity to the sides as well as to keep moving cooler air thru, around the cooler and out of the case.

 

Yes, this is precisely what I was saying. The standard 140 fan supplied by Noctua, is wider than the heat sink. This is deliberate on Noctua's part. Designed this way so that some air bypasses the heat sink and cools the motherboard. You have now replaced that 140 fan with a bigger 150 fan, and in addition added a second 150 fan to the heat sink. Clearly, where the extra diameter of the 150 fans overlap, they are directing air outside of the heat sink [rather than through the heat sink]. Thus the extra diameter provided by the bigger fans is not contributing to CPU cooling at all, only motherboard and case cooling. So the bigger fans can't cool your CPU any more than the standard fans.

 

I suspect that the slight improvement in CPU cooling you have experienced is more to do with the fact that you have cut out the rear fan grill, which would have helped considerably if it was the restricting kind. Plus removed the rear case fan, and fitted a replacement fan directly to the heat sink.  In any case, we are not talking about big changes in temperature.

 

What about the fan at the bottom of your case, was this added prior to the mod or after, so not a variable?

 

 

Great stuff that your system is now quieter and that you're happy, but what I will say is that I can't hear my D14 anyway, above the noise from all of the other fans in my enclosure. Enclosure fans are running at low RPM, and D14 fans are full speed. The D14 has never been noisy, so no need to make it quieter.

 

You have a prodigious number of case fans there. I count 5, Could you hear the D14 above the noise of those case fans, before your mod?

 

 

 

It works.

 

 

 

But I don't think it works by a significant amount.

 

I'm not deriding your mod, I think it's a great mod if you are happy, I'm just trying to determine if it's worth it for me and anyone else with a D14, in terms of temp reduction and noise reduction. And to do that I need to delve into the details. No offence intended.

[Appliance 51]

What about the flying, chaps?  :lol:

[martin-w 3,235]

:smile:

 

Glad the comma's there, I was going to say, I haven't seen any flying chaps lately.

[Guest]

Oh my, this thread is still going ... just curious, but I assume you folks are overclocking using ROG panel?

[Boomer 1,619]

Dudes, what is the verdict?

 

I currently have a air coller master V8 for my o/c i920...

 

BUT am moving to a 4670K this week.

 

I am looking at a AIO water cooler for better temps to o/c, maybe a Corsair H60/80, nothing too expensive.

 

Is it reasonable to expect better temps with AIO versus air?

[Guest]

I've pre-order one of these: http://www.newegg.com/Product/Product.aspx?Item=N82E16813132053

 

ROG OC panel is a great way to really get the most out of one's hardware ... it reports error conditions and really makes it that much easier to obtain high OC with great stability.

 

Disclaimer: I'm doing this for fun -- I have no expectations or assumptions that doing all this work to get 1-5 fps extra out of FSX or P3D v2 is in any way worth the time and money investment.  For me, it's the fun of trying ... maybe even failing and burning up some components in the process.  IMHO, the motherboard has always been the most important single component for any PC build -- this is the base from which I build around.

 

Rob

[martin-w 3,235]

Dudes, what is the verdict?

 

I currently have a air coller master V8 for my o/c i920...

 

BUT am moving to a 4670K this week.

 

I am looking at a AIO water cooler for better temps to o/c, maybe a Corsair H60/80, nothing too expensive.

 

Is it reasonable to expect better temps with AIO versus air?

 

 

Well no, it depends which air coolers and which AIO coolers you are talking about. Just because it's water, doesn't necessarily mean it's better. 

 

The D14, Silver Arrow, the handful of top air coolers, beat the H60/80 with ease. They beat the H100i as well.

 

If you want an AIO cooler that beats the top air coolers, then you are  looking for more surface area.

 

The H110, Kraken etc. The H110 is currently the best, when you eliminate the fan variable.

 

 

http://www.guru3d.co..._review,13.html

 

H100i 76 degrees

Kraken X60 70 degrees [Noisy higher RPM fans]

Corsair H110 72 degrees [Very quiet]

NH-D14 72 degrees [Very quiet]

 

 

Eliminating the fan variable, from the previous page...

 

 

What we almost never see is a comparison with "the same fans". What should concern us in the efficiency of the heat sink itself, without the fan variable.

 

When we do that, we truly see which are the best coolers. For example, if you look at the link below, you can see that when the same fans are used, rather than the NZXT high RPM noisy fans, that the Corsair H110 beats the X60 by 5 degrees.

 

https://docs.google....5GSFZNRmc#gid=0

 

[PigsInSpace 21]

Standard stuff when we test. Prior to mod run a stress test, full load, full RPM, determine average temps across the cores. Make a note of ambient temp. Keep all variables constant. Make a note of the delta T. Run the same test again after the mod. Time isn't critical as long as it's the same for each test, and provides sufficient time to reach max temp.

 

If we don't do this, and keep accurate notes, we have no idea.

 

 

 

1-2 degrees is not significant statistically, as it's within the margin of error we would expect when testing any cooler. 2-3 degrees is minimal. So yes, pretty much what I expected, not much of an improvement in terms of temp compared to the D14 standard set up.

 

 

 

 

 

No, it's certainly not oval. :smile:

 

 

 

Yes, this is precisely what I was saying. The standard 140 fan supplied by Noctua, is wider than the heat sink. This is deliberate on Noctua's part. Designed this way so that some air bypasses the heat sink and cools the motherboard. You have now replaced that 140 fan with a bigger 150 fan, and in addition added a second 150 fan to the heat sink. Clearly, where the extra diameter of the 150 fans overlap, they are directing air outside of the heat sink [rather than through the heat sink]. Thus the extra diameter provided by the bigger fans is not contributing to CPU cooling at all, only motherboard and case cooling. So the bigger fans can't cool your CPU any more than the standard fans.

 

I suspect that the slight improvement in CPU cooling you have experienced is more to do with the fact that you have cut out the rear fan grill, which would have helped considerably if it was the restricting kind. Plus removed the rear case fan, and fitted a replacement fan directly to the heat sink.  In any case, we are not talking about big changes in temperature.

 

What about the fan at the bottom of your case, was this added prior to the mod or after, so not a variable?

 

 

Great stuff that your system is now quieter and that you're happy, but what I will say is that I can't hear my D14 anyway, above the noise from all of the other fans in my enclosure. Enclosure fans are running at low RPM, and D14 fans are full speed. The D14 has never been noisy, so no need to make it quieter.

 

You have a prodigious number of case fans there. I count 5, Could you hear the D14 above the noise of those case fans, before your mod?

 

 

 

 

 

But I don't think it works by a significant amount.

 

I'm not deriding your mod, I think it's a great mod if you are happy, I'm just trying to determine if it's worth it for me and anyone else with a D14, in terms of temp reduction and noise reduction. And to do that I need to delve into the details. No offence intended.

 

These systems are Flow-tmp bench tested (as used at ForzenCPU) with a whole slew of 24 hour burn-in runs, its as accurate as it gets.

 

No, its not about the missing grill (one of the first mods you should make, you want airflow or not?) its about the added fan minimizing drag and maximizing airflow thru the cooler.

 

I provided honest real world results not some vague wishful unproven crap, I'm not here to waste your time, please don't waste mine.

2- 3 Celsius repeatable difference on accurate consistent testing is a real gain in extreme testing, compares very well with other efforts you could spend time money on and given it is just an added fan running at the chosen speed of 750 rpm and don't miss the point - its not about breaking records but it is about being 2/7 stable at high overclock in a quiet and efficient way (notice 4.4GHZ Haswell - 16GB decently fast ram running thru the IMC non de-lidded effortlessly with room to spare).

 

Case fans? Noise?

 

You must have missed the post were I explained that ALL the case fans are dead silent at 500 RPM. 

 

And no, the 140mm fan does not cover up the D14, it in fact is non effective on about 15% the surface area missed at the corners, the 150mm slightly improves on that.

 

Overclocked, Cool and Quiet. No muss no fuss.

 

Cheers.

[Guest]

Here are some pics of my primary work/simulator PC ... H100 (2 x 120mm fans) with 140mm rear fan, 230mm front fan, 230mm side fan, and I added 140mm to draw air out behind the CPU mount ... the motherboard carrier already had a large square cut out for heat exhaust so I made 140 mm hole is the case side panel and added a 140mm fan to help draw that heat out.

 

Fan speeds are very low (600 to 1100 rpm) even under Prime95 (pic is for example, I usually run prime for 8 hours after any OC changes) loads ... CPU will hit about 154 F (around 67-68C).  Rock solid stable and I stress it much more than FSX does, C4D renders and Adobe CC renders.  <21 db

 

It's running with HT on (6/6) 3960X @ 4.8 Ghz with 2133Mhz DDR3  (32GB RAM frequeny is 1066Mhz with 9-11-11-31-260-2T) 1.65v quad channel.  Memory bandwidth and latency are extremely good.

 

For a basic water cooling setup and a proper fan placement, this system has been rock solid and very quiet.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Rob

[wizzards 99]

ihe photos by pigs in space show what i dislike about that cooling set up ,which I use,it is  to bloody big ,when I want to clean the fans/fins change memory etc its a night mare for me ,the cooling however is excellent keeps my i2600k at 4.5 nice and cool which is it,s job ,but I feel that a Swiftech H20-220 is calling ,does anyone use this cooler,is this a good choice ?

peter

[martin-w 3,235]

I provided honest real world results not some vague wishful unproven crap, I'm not here to waste your time, please don't waste mine.

 

 

No need to be upset, there's no need to be defensive, I have no wish to aggravate you. I'm just trying to determine if your mod is worthwhile for me to try, or others with the D14.

 

 

 

 

2- 3 Celsius repeatable difference on accurate consistent testing is a real gain in extreme testing,

 

 

it is about being 2/7 stable at high overclock in a quiet and efficient way (notice 4.4GHZ Haswell - 16GB decently fast ram running thru the IMC non de-lidded effortlessly with room to spare).

 

These systems are Flow-tmp bench tested (as used at ForzenCPU) with a whole slew of 24 hour burn-in runs,

 

 

If it's worth it for you, buying two new expensive Noctua fans, and going to the trouble to cut out your rear fan grill, for a mere 2-3 degrees, then great stuff. For me though, it would be pointless. The D14 in standard form, is the quietest high end cooler I have ever owned, I have no need to make it even quieter, I can't hear it anyway. And my overclock certainly isn't on a thermal knife edge, and requiring a mere 2-3 degrees temperature drop. 

 

Basically, I don't feel you needed that minor 2-3 degrees drop in temperature. And you certainly didn't when you consider that in 24/7 use, your temps will be nowhere near the temperature experienced during a  "slew of 24 hour burn-in runs". In 24/7 use, running FSX or whatever else you do, you will have loads of temperature latitude.

 

 

 

 its about the added fan minimizing drag and maximizing airflow thru the cooler.

 

 

 

 

Well, it's not exactly an added fan. You have taken away your rear case fan, and added a fan in contact with the heat sink. So effectively you have simply moved the fan closer to the heat sink. If you read my post, I did mention that one of the factors responsible for your slight temperature drop, was placing a fan in contact with the cooler.

 

At the end of the day, all that matters is that you are happy. I'm glad you are, but for me, your modification would not be worth it. Apologies if I annoyed you by being argumentative and challenging your assumptions, but that's how we learn.