Cooling in summertime & heatsink list equality question

[Silence & Science]

Hey guys! As I’m new here let me start by saying: thanks for existing! It’s somehow nearly impossible to find consistent quality independent noise-related info on pc components, so SPCR really seems like the one rare gem in this aspect.
Seen as how I’m a rookie on these forums ill post this here, though my questions are mostly cooling related.

The goal: A silent PC with some umpf that I can use to study/work, listen to music and for fun time. Usage will vary from lazing around on youtube to gaming/multitasking with a gazillion chrome, pdf, word and excel windows open simultaneously and occasionally video/audio editing or some statistical analyses.
My background ambient noise level comes in at around 24~26 dBA (SPL meter), but it may well be lower as that is also around the SPL meters minimum value, but it's pretty quiet especially at nights. Ambient temps vary from 17-37 degrees (occasionally even 40ish degrees inside) Celsius due to non-AC and poor ventilation.
I was thinking of getting something along the lines of:

- Skylake 6700k (or potentially a 6600K but I’m leaning towards the i7 for HT)
- 16GB DDR4 2400 RAM
- GTX960 or R9 285 or the likes
- Samsung 850 EVO SSD
- Some random silent HDD or my current 7200 rpm one for storage
- Fractal Design R5 case
- Some silent PSU
I’m not sure if I’ll need to OC the i7 (I probably would on the i5 eventually though), but I probably won’t as long as it suffices.
Having looked through the SPCR website I do still have a few cooling related questions:

1. I may be missing something here but: The Megahalems lists as producing a 44 degree temperature rise at 11~12 dBA in the heatsink comparison table, in the actual review this is much less across the board, so where do the 44 degree marks etc. in the recommended list come from?

2. The Megahalems and NH-D14 are both tested on a different platform (each) with nexus 120mm fans if I’m correct. This is different from the others, and the NH-D14 is tested with 2 fans and the Megahalems with 1. It appears that both the NH-D15 and NH-U14S tend to beat both of them in raw cooling/noise performance in just about any other review, so I was wondering how do I compare these two coolers to the rest?

3. The Scythe Kotetsu is awesome value for money. However I do wonder if it will remain quiet enough if in summer ambient temp. is around 37~38 C° in a poorly ventilated, non-airconditioned room (it’s a rental, what can I say…) when not OCing. Ideally it has to be kept at 7v or below, though 9v is still doable. Several reviews testing the Kotetsu found it reaches 65C°ish full-out with a 4770k or similar CPU when testing at ambient temps of 20ish degrees, at ambient 37 the CPU would then be around 82 degrees. The 38 C° rise at SPCR gives a similar indication, if I can just add that to ambient I’d end up with around 76 C°. I think this may be a bit much for silent computing?
So I’m wondering if the Kotetsu will cope whilst remaining silent/quiet enough, or whether it’s maybe a better idea to go with a bit more heavy-duty cooler? Say a Prolimatech Genesis/Megahalems, Thermalright True Spirit 140 Power or a Noctua variant?

4. I’ve seen several people coming to the conclusion that some fans tend to either “click” at PWM control or get tonal/hummy when placed on a heatsink. Is there any way to tell whether this will happen with certain fans, for e.g. if higher static pressure equals more tonal noise or is this truly specificly fan-dependant? Or is it that the motherboard control plays a role in clicking noises?

Any thoughts/help on this would be appreciated!


I'll also sketch my current situation for comic effect: (feel free to ignore this)
So I am currently sitting next to a 5,5 year old AMD Phenom II X4 955 PC with stock cooler and dried-out thermal paste running XP. This was ironically offered as one of the more “silent” options by the vendor, probably because Cool ‘n Quiet was implemented or something, though it was always quite loud under any load from day 1.
Needless to say it’s currently doing an awesome job keeping my CPU at 64C°ish at around 4000 rpm in BIOS only, which is quite noisy. I just closed a Chrome window, CPU jumped to 72C° (chrome + word open). This gets worse in summer as temperatures go up to 37 C° (last summer I think we hit 40 even). Seen as we also have no airconditioning the PC does an awesome impression of a jet engine at times, which I found can actually be heard outside through the closed door and windows 1 floor down when it’s running idle. Which is of course a weird thing to explain to your visitors when dining outside during a nice summers eve… So rather than to patch this up and upgrade I figured I’d just built something from scratch that is actually silent and up to date so it will last a while and so I can focus... On, to Silence!!

[quest_for_silence]

17-37 degrees (occasionally even 40ish degrees inside) Celcius

Given your moniker, Celsius please (celcius is a misspelling in any language).

1. I may be missing something here but: The Megahalems lists as producing a 44 degree temperature rise at 11~12 dBA in the heatsink comparison table, in the actual review this is much less across the board, so where do the 44 degree marks etc. in the recommended list come from?

From some nearly silent (pun not intended) retest at some platform/methodology update.

2. The Megahalems and NH-D14 are both tested on a different platform (each) with nexus 120mm fans if I’m correct. This is different from the others, and the NH-D14 is tested with 2 fans and the Megahalems with 1. It appears that both the NH-D15 and NH-U14S tend to beat both of them in raw cooling/noise performance in just about any other review, so I was wondering how do I compare these two coolers to the rest?

You have to link to your references, otherwise we cannot comment with ease.

Broadly speaking: 1) SPCR chart cannot be compared to other charts; 2) IIRC ranking is based upon the highest SPL column, that's why you see the D14 before its updated sibling; 3) IIRC any heatsink in the reference chart has been tested on the same platform (that's why there's the D14 with 2 Nexus).

3. The Scythe Kotetsu is awesome value for money. However I do wonder if it will remain quiet enough if in summer ambient temp.

It mostly depends on actual load and how you will set the fan curve (mainly, the relevant hysteresis): broadly speaking, probably yes.

Is there any way to tell whether this will happen with certain fans, for e.g. if higher static pressure equals more tonal noise or is this truly specificly fan-dependant?

It's fan dependant.

Or is it that the motherboard control plays a role in clicking noises?

No, it's the fan motor.

[CA_Steve]

fyi - if you edit your first few posts, it throws them back into the moderator queue

[Silence & Science]

Given your moniker, Celsius please (celcius is a misspelling in any language).

Oh snap that's an emberassing spelling mistake Well what can I say, it was really late...but you're totally right of course! Edited it & thanks!

From some nearly silent (pun not intended) retest at some platform/methodology update.

Allright that makes sense! I was just confused as I didn't encounter this in the actual review.

You have to link to your references, otherwise we cannot comment with ease.

Broadly speaking: 1) SPCR chart cannot be compared to other charts; 2) IIRC ranking is based upon the highest SPL column, that's why you see the D14 before its updated sibling; 3) IIRC any heatsink in the reference chart has been tested on the same platform (that's why there's the D14 with 2 Nexus).

Ah yes that in combination with your previous answer renders this question obsolete, thanks!

It mostly depends on actual load and how you will set the fan curve (mainly, the relevant hysteresis): broadly speaking, probably yes.

That's a good point, I suppose the difference in CPU temperature is negligible after 9V anyway. But under those conditions (significant load, higher ambient temps) wouldn't a silence oriented fan curve reduce the airflow inside the case in such a way that the non-CPU components heat up more and thus cause more noise, for e.g. via the GPU or the case fans?

It's fan dependant.

That's a shame, but thanks!

No, it's the fan motor.

Ok thanks!

fyi - if you edit your first few posts, it throws them back into the moderator queue

Haha thanks, I noticed, and for such a futile edit too, sorry moderators!

[quest_for_silence]

Edited it & thanks!

Don't mention it, thank you too.

That's a good point, I suppose the difference in CPU temperature is negligible after 9V anyway. But under those conditions (significant load, higher ambient temps) wouldn't a silence oriented fan curve reduce the airflow inside the case in such a way that the non-CPU components heat up more and thus cause more noise, for e.g. via the GPU or the case fans?

Well, more probably that not, no: with a side-cooler you cool about just the CPU. Sometimes you might affect, slightly, the mobo VRM temp but again, more probably that not, that may be of some importance for linpack/prime avx kind of loads.