Overclock.net › Forums › Cooling › Air Cooling › Cooler upgrade
New Posts  All Forums:Forum Nav:

Cooler upgrade - Page 2

post #11 of 23
Quote:
Originally Posted by alltheGHz View Post

Hey man, welcome to OCN! Your first post has more responses than my build log, mines been up for twice as long as yours, gg m8!

Anyways, For the amount you're gonna spend on the coolers, spend a bit more and get the corsair h60. In a push pull configuration, I'm able to achieve ~30 degrees celcius on my FX6300 OCd. Plus, you wont have to worry about clearance, blocking, ect.

Well, there is the whole issue with the H60 being louder, more expensive, and not performing as well as the above mentioned coolers:

post #12 of 23
Here we go again.

Simple fact: Top air coolers are as good or better than top CLC for less money, run quieter, are more dependable, and the only thing that can go wrong is the fan/s .. cheap and easy fix. The most common problem with CLCs is the pump. When pump fails your system has no cooler and can't be used until the complete CLC is replaced. .. not cheap or easy.

@ WideVector
It's the inaccuracy of most testing. thumb.gif

The single biggest problem with cooler testing and reviews is very few are tested using the cooler intake air temperature as the baseline. Most reviewers use the room temperature as a baseline, not the actual temperature of air going into the cooler .. and they are not the same!

Using room temp is like looking at the thermometer in your bedroom to find out how warm your kitchen is. They might be the same .. but probably are not.

Every change in air temperature going into cooler causes a similar change to CPU temperature.

Even when testing on a open test bench the cooler intake air temp is often not the same as the room. The heated air coming off of GPU, CPU, PSU, even a human body changes the air temp. And the closer these are to cooler, the more the change. I have seen 8-15c difference on open bench! (With down-flow coolers). Testing in a case without monitoring cooler intake air temperature is at best just plain stupid .. at worst knowingly deceptive .. smoke and mirrors because they don't want accurate results. Using a case on the pretext of simulating "real world use"?? Really?? The only "real world" invoked is the testers system in their environment. Each of our "real worlds" is different than the tester's .. I would be surprised if 1 in 1000 is close to being the same.

Same applies to most testing results comparing large numbers of coolers. It's not just the air temperature that determines how well the air absorbs the heat from cooler, the air pressure, humidity, etc. all effect it. We have to allow at the minimum of 2c margin of error (+1c / -1c) .. and this can realistically be 4c margin of error when we take variances in production quality of cooler, of fans, TIM seat, etc. into account

It's a conspiracy! wheee.gif

And just because I'm paranoid doesn't mean they are not watching me. whistle.gif

** Climbs down off of soapbox **
** Takes a deep breath **

That said, there are some small coolers that do a very good job. Much of this is the quality of construction.


Most motherboards will accept 140mm fan coolers .. Coolers 140mm wide. But there are a few that will not. Thermalright has several coolers with offset bases for this reason.

Edited by doyll - 3/25/15 at 5:19am
post #13 of 23
Quote:
Originally Posted by WideVector View Post

Thanks for those links. I will make my way through that thread over time. That review though, is one I haven't seen so thanks for that.

Something caught my eye though. In a couple of those tests smaller coolers, with less heatpipes, like the tc12dx and u12s match or beat the TS140 and other similarly large coolers. How is that possible? Surely a 4820k at 4.3GHz produces enough heat to allow the larger coolers to pull away. I've seen results like that in other reviews and start to question the reliability of those reviews because on paper a cooler with more heatpipes and more fin surface area should cool better.

I will say though, if that review is accurate the macho 120 appears all over the place so I think that one is out.

In addition to doyll's facts regarding cooler intake temps.....

Looking good on paper does not tell the whole story of a tower. If that were the case, life would be easy and we could simply go by size. What isn't on the paper is the design and quality if the heatpipes themselves, the copper used (cool sheet on copper composition and conductivity - http://olinbrass.com/sites/default/files/downloads/Olin-Brass-Copper-Alloy-C110-Data-Sheet.pdf) in the contact block, the thickness of the nickel coating, solder and welding techniques used, fin density and material composition, fin design, etc, etc, etc. These all have affects on performance. Companies like Noctua, Phanteks, Cryorig, Thermalright excel in these areas - it is a huge reason their coolers outperform knockoffs that look almost identical to them.

More simply put - you could have the biggest tower available with eight 8mm heat pipes, and if the soldering to the fins is very poor (or pick another condition from above)....it will have very poor performance.
Edited by ciarlatano - 3/25/15 at 4:25am
post #14 of 23
Quote:
Originally Posted by doyll View Post

Here we go again.

Simple fact: Top air coolers are as good or better than top CLC for less money, run quieter, are more dependable, and the only thing that can go wrong is the fan/s .. cheap and easy fix. The most common problem with CLCs is the pump. When pump fails your system has no cooler and can't be used until the complete CLC is replaced. .. not cheap or easy.

Those results were taken from the Shadow Rock Slim review. I haven't done comparisons to the H60 since then - I figured after watching it get beat up by $35 air coolers time and time again.... But a little more perspective using the Shadow Rock Slim as a control:




Using the SRS as a control comparison....the H7 and Pure Rock are half the price of the H60, and offer better performance with far less noise.....oh, and that whole reliability thing.

One thing that never made the SRS review - you'll note that the <40 dB testing never made it into the review. I figured it was a moot point since all the comparison units were well under that mark, excepting the H60 which was already getting beat up making all its noise. But, I did actually test the SRS against the H100i at <40 dB. I didn't publish the results for fear of massive heart failures on the Lido Deck of the sailboat.
post #15 of 23
Quote:
Originally Posted by ciarlatano View Post

Those results were taken from the Shadow Rock Slim review. I haven't done comparisons to the H60 since then - I figured after watching it get beat up by $35 air coolers time and time again.... But a little more perspective using the Shadow Rock Slim as a control:
Warning: Spoiler! (Click to show)


Using the SRS as a control comparison....the H7 and Pure Rock are half the price of the H60, and offer better performance with far less noise.....oh, and that whole reliability thing.

One thing that never made the SRS review - you'll note that the <40 dB testing never made it into the review. I figured it was a moot point since all the comparison units were well under that mark, excepting the H60 which was already getting beat up making all its noise. But, I did actually test the SRS against the H100i at <40 dB. I didn't publish the results for fear of massive heart failures on the Lido Deck of the sailboat.
LOL

I was referring to alltheGHz 's post thumb.gif

As soon as I get time I'm going to comparison test H5 and a few others against the TRUE Spirit 140 coolers. Last time I comparison tested TRUE Spirit 140 it was significantly cooler than 120mm fan coolers.
Edited by doyll - 3/25/15 at 5:28am
post #16 of 23
Quote:
Originally Posted by doyll View Post

LOL

I was referring to alltheGHz 's post thumb.gif

As soon as I get time I'm going to comparison test H5 and a few others against the TRUE Spirit 140 coolers. Last time I comparison tested TRUE Spirit 140 it was significantly cooler than 120mm fan coolers.

No worries, I knew that. I was simply adding to the section of your post that I quoted.
post #17 of 23
Thread Starter 
Quote:
Originally Posted by alltheGHz View Post

Anyways, For the amount you're gonna spend on the coolers, spend a bit more and get the corsair h60. In a push pull configuration, I'm able to achieve ~30 degrees celcius on my FX6300 OCd. Plus, you wont have to worry about clearance, blocking, ect.

Thanks for the suggestion but I am simply not interested in any of those AIO water coolers.
Quote:
Originally Posted by doyll View Post

It's the inaccuracy of most testing.
If that's the case then are there any reviews sites that you trust more than others?

I understand what you're saying about room temp vs air intake temp, and please believe me when I say I'm not trying to doubt you, but, wouldn't the room temperature be a fairly good indicator of what the air intake temperature is? I can completely understand a few degrees difference, but assuming reviewers used the same case, or open bench in the same location for each cooler with the same intake fans I am struggling to see how the difference could be more than a few degrees.

I'd be interested to read any articles on the matter or your review process where you saw that large difference. Again, not trying to doubt you just looking for more information on the subject.
Quote:
Originally Posted by doyll View Post

Most motherboards will accept 140mm fan coolers .. Coolers 140mm wide. But there are a few that will not. Thermalright has several coolers with offset bases for this reason.
If that's true then what would some of the better coolers be that fit within 140mm width? Or more specifically 70mm center of CPU to first PCI-e slot.

@ ciarlatano
Thank you for that as well. I always knew quality was a big part of a coolers performance but always just thought that was related to fin thickness, the flatness of the base, and the quality of the soldering. Never occurred to me that difference manufacturers could be using different alloys, thanks for that link by the way. Your comment about nickel plating made me look up the thermal conductivity of it. According to Wikipedia it's less than half that of Aluminum and about a fourth of Copper, did not expect it to be that much worse.

Here's the link for those interested: http://en.wikipedia.org/wiki/List_of_thermal_conductivities
post #18 of 23
Quote:
Originally Posted by WideVector View Post

Thanks for the suggestion but I am simply not interested in any of those AIO water coolers.
If that's the case then are there any reviews sites that you trust more than others?

ciarlatano and ehume do good testing. Best is to check and see what air temp they are using when testing.

I understand what you're saying about room temp vs air intake temp, and please believe me when I say I'm not trying to doubt you, but, wouldn't the room temperature be a fairly good indicator of what the air intake temperature is? I can completely understand a few degrees difference, but assuming reviewers used the same case, or open bench in the same location for each cooler with the same intake fans I am struggling to see how the difference could be more than a few degrees.

Think of your house with central heating as our room and your living room as your computer, with a fireplace (CPU and wood stove (GPU). If you have a thermometer in your dining room it will read about the same as one in the living room .. until you light the stove and fireplace. Then the living room will become much warmer.

Don't believe me. get a cheap indoor/outdoor digital thermometer with a wired sensor like I show in case cooling link below and check it out for yourself. You will find out just what I mean.


I'd be interested to read any articles on the matter or your review process where you saw that large difference. Again, not trying to doubt you just looking for more information on the subject.
If that's true then what would some of the better coolers be that fit within 140mm width? Or more specifically 70mm center of CPU to first PCI-e slot.

Post #18 show room 22c with cooler intake 30c .. and a 140mm fan blowing air across top of cooler on test station.
http://www.overclock.net/t/1347211/thermalright-axp-100-showed-up-today/0_20


Quote:
Test Methodology, Hardware & Software

All coolers were tested in an open case lying flat for ease of fan swapping and intake temperature monitoring. A Thermalright TY-147 (1200rpm) was positioned over the chipset to keep the motherboard and memory cool. Ambient room temperature was not measured; instead, a more specific and accurate measurement of intake temperature was used when calculating final results (thanks to doyll for this suggestion). The image below shows how I incorporated this approach into my testing, with the thermistor from a small electronic thermometer placed directly in the path of the intake fan.

I found that the intake temperature almost always differed from the ambient room temperature and, perhaps more significantly, it also varied a lot between tests, depending on the fan speed used and the positioning of the thermistor. With higher fan speeds the intake temperature fluctuated significantly because the exhaust air from the cooler was sometimes circulated back into the path of the intake fan; and with lower fan speeds the intake temperature was generally higher because there was no air current to remove the heat being dumped by the motherboard and memory. All this was taken into account when calculating the final results and intake temperature was closely monitored throughout the process.
http://forums.bit-tech.net/showthread.php?t=269339

One of the links in my sig is to a chart of many top cooler sizes and clearances.

Edited by doyll - 3/25/15 at 4:54pm
post #19 of 23
Thread Starter 
Quote:
ciarlatano and ehume do good testing. Best is to check and see what air temp they are using when testing.
Good to know, I will look out for reviews by them as well.

Quote:
Think of your house with central heating as our room and your living room as your computer, with a fireplace (CPU and wood stove (GPU). If you have a thermometer in your dining room it will read about the same as one in the living room .. until you light the stove and fireplace. Then the living room will become much warmer.
That analogy makes a lot of sense, thanks.

Quote:
Post #18 show room 22c with cooler intake 30c .. and a 140mm fan blowing air across top of cooler on test station.
http://www.overclock.net/t/1347211/thermalright-axp-100-showed-up-today/0_20
Interesting. Glad to see you reviewed that cooler, I just bought one for my HTPC, haven't had a chance to install it yet though.

Quote:
...the exhaust air from the cooler was sometimes circulated back into the path of the intake fan;
That is a very good point, and something that complete slipped my mind, especially with non-reference GPU coolers that dump air back into the case. Even in a pseudo wind-tunnel that is the FT02 that could have a big effect. Thanks for that link, bit-tech has been one my go-tos for a while. And thanks for explaining that more and providing some examples. I know you've said just about everything you've said here many times, and I appreciate you going into more detail.

Quote:
One of the links in my sig is to a chart of many top cooler sizes and clearances.
I have looked at that chart of yours before but I was in a different mind set about what I wanted and cooler sizes. I will definitely take another look at it.
post #20 of 23
Thread Starter 
Did some more searching and found a review that includes all the coolers I've looked at; Macho 120, Deepcool Neptwin, and a lot, if not all, of other popular coolers, including the True Spirit 140 recommended by doyll. It's russian website so google translate might be needed. This is one of the better reviews I've seen, and love that it includes so many coolers.

http://www.3dnews.ru/818946/page-2.html

Overall temperatures with noise levels



This one shows the maximum overclock achieved by each cooler with the respective temperatures and noise levels.



After looking at these I can see why doyll recommends the True Spirit 140 so much. It's one of the best temperature wise, and at low fan speeds, one of the quietest while still giving very good temperatures. Once you include the price of $50 shipped from Amazon, also one of the cheapest. I didn't originally want a cooler this large, but looking at these, and other results, it's a no brainer and I am completely sold on it.
Edited by WideVector - 3/28/15 at 4:38pm
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Air Cooling
Overclock.net › Forums › Cooling › Air Cooling › Cooler upgrade