vertical motherboard mount w/cooler

[The Original Ralph]

this is my first build with a case mounting the mobo vertically and i'm trying to decide between liquid or air cooling.

i'm adverse to using a liquid cooling system, but have questions about long term effect of the weight of an air cooler on the motherboard. With the liquid cooler, i could fab a fiberglass drainage sheet to carry any leakage to a non-critical area but would rather not.

i have worked with fiberglass / epoxies on a professional level, and epoxies will soften with heat. With the heat the cpu is generating, in this case a 140 watt cpu, my concern is the weight of a 1 lb (noctua NH-U14S) cooler that is 160mm tall - the moment arm of that wgt is roughly 80mm, which still means a 3 - 3.5" leverage of that weight on the mobo.

Do people experience the motherboard sagging from that wgt & heat over long term?

 

[mcnumpty23]

not aware of it causing any major issues using a large air cooler millions of people do it

 

[geofelt]

What cpu do you need to cool?
A good tower type cooler like a noctua NH-U12s will cool most anything and not be overly heavy.
My canned rant on liquid cooling:
------------------------start of rant-------------------
You buy a liquid cooler to be able to extract an extra multiplier or two out of your OC.
How much do you really need?
I do not much like all in one liquid coolers when a good air cooler like a Noctua or phanteks can do the job just as well.
A liquid cooler will be expensive, noisy, less reliable, and will not cool any better
in a well ventilated case.
Liquid cooling is really air cooling, it just puts the heat exchange in a different place.
The orientation of the radiator will cause a problem.
If you orient it to take in cool air from the outside, you will cool the cpu better, but the hot air then circulates inside the case heating up the graphics card and motherboard.
If you orient it to exhaust(which I think is better) , then your cpu cooling will be less effective because it uses pre heated case air.
And... I have read too many tales of woe when a liquid cooler leaks.
google "H100 leak"
-----------------------end of rant--------------------------

Your pc will be quieter, more reliable, and will be cooled equally well with a decent air cooler.

 

[The Original Ralph]

THE intel i7-5960x and i do plan on OCing some - somewhere in the area of 4.2-4.3, reviewers seem to all settle on 4.5 as the highest stable OCing level

i like the U12S, and am already using one right now on my i7-4790 - but the 4790 has a TDP of 84 watts, and the U12S keeps it comfortably in mid 60C when rendering video files - the main program i use is a core hog and keeps the 4790 at 98-100% load on all four cores. With 140 watts, i don't see the U12S keeping temps anywhere below 75C. The U14S would add a couple of heatpipes and a little margin cooling wise.

While i agree with most of what you say in your rant, one thing you haven't taken into consideration - water is a better absorber of heat or has a higher co-efficent of thermal conductivity - basically it'll acquire the heat faster / better than air

but oddly, when i look at comparisons of coolers, the liquid coolers don't seem to achieve but 3-4 degrees cooler temps, which to me is not worth the risk of a liquid cooling system, especially when the right leak can mean wiping out a small fortune in mobo/graphics card/cpu investment.

Like i said earlier i worked with fiberglass for a long while - what i don't understand is why we aren't seeing carbon fiber cooling devices - fyi, carbon fiber (or even just carbon) has the highest rate of thermal conductivity. That's why a few of the thermal paste use diamond dust in their paste - diamonds are carbon. It conducts heat almost as fast as copper conducts electricity, i kid not. For the hey of it, if you can get your hands on a small pc of carbon fiber cloth (before it's been laid up with epoxy resin) hang it from a vice, and with a common propane torch try to get a spot on it glowing red. IF you're real careful and hold the torch at the exact distance from the cloth, you'll finally get a spot the size of a 25cent coin (a quarter). The instant you pull the torch away, it goes back black and 1.5 seconds later you can actually touch the spot you had glowing cherry red. A real life example - we mfgr firearm suppressors (silencers). One model that we offer, that is normally assembled in an aluminum tube, we fabricated some with carbon fiber tube bodies. After 450 rounds thru the alum suppressor on a 94F day, temp on the can was 174F - too hot to touch. The CF can, 114F. THe aluminum can took 14-15 minutes to cool down enough to put back in it's ballastic nylon pouch while the CF can took 1.5 minutes - it was drawing the heat out of the aluminum internal components and transferring it to the air.

I keep thinking about fabbing a 1/8" thick pad to put between the cpu and the cooler's heatsink, let it extend out past the heatsink maybe 1/2" on all sides, and see what kind of improvement i could achieve.

but back to your answer - do you really think the U12S would handle 140 watts? I know Noctua's site indicate it's rated for 140 watts, but it's seems to be close to it's limit with the 84 watts of the 4790. And i have that case pressurized with two extra fans blowing air in.

 

[amtseung]

Really, really heavy coolers (Scythe Ninja Rev4 and those gaudy, overpriced cooler master things come to mind) will eventually cause sagging and may even slowly warp your motherboard over years, but coolers that massive can probably be run passively. Liquid cooling is cool and all, but it's a helluva routine to regularly maintenance the loop, let alone change out the components underneath. AIO coolers are always an option, and a nice middle ground for somewhat maintenance-free water cooling solution. You still have to clean all the crap out of the radiators once in a while.

Will the U12S handle a 5960x? Probably. If they can handle lightly overclocked FX8350's that usually run estimated 175W+ TDP, yeah, probably. No guarantees. How about the NH-D15 or D15S or something off the Cryorig lineup?

 

[The Original Ralph]

well, the consensus seems to be the mobo sagging or warping is not an issue

as to the D15, we're starting to get into serious tonnage there - noctua specs the wgt at 1320 grams or 2.910 lbs
the U14S is 935 grams or just over 2 lbs. The D15S comes in, surprisingly at 1150 grams or 2.5x lbs. But both D15s would be a pain to clean properly.

the U14S can also be boosted some with a 2nd fan, and if temps are still an issue, i'll look around for a nitrogen AIO system - someone has to be offering one, somewhere.

tks for the response

 

[amtseung]

well, the consensus seems to be the mobo sagging or warping is not an issue

as to the D15, we're starting to get into serious tonnage there - noctua specs the wgt at 1320 grams or 2.910 lbs
the U14S is 935 grams or just over 2 lbs. The D15S comes in, surprisingly at 1150 grams or 2.5x lbs. But both D15s would be a pain to clean properly.

the U14S can also be boosted some with a 2nd fan, and if temps are still an issue, i'll look around for a nitrogen AIO system - someone has to be offering one, somewhere.

tks for the response

I've heard of people propping up their sagging D14's and Prolimatech Mega-WTF-behemoths with fishing line, tying it to the top of the case as a support, or fabbing a support rest directly into the side panel directly opposite the heatsink tower. Heat sinks are far easier to clean than the higher fin density radiators used for water cooling. At least you can use a leaf blower and not risk damaging anything.

Strapping a second fan for a push-pull config on a single stack heatsink like the U14S (and U12S and Hyper 212 and all of these) isn't as effective as you'd think. It'll help a bit, but the extra weight, noise, power consumption, and upfront cost don't justify the 2-3C difference, especially with the cost of a Noctua fan. The dual tower heat sinks definitely benefit from a second fan though. Heat soak is heat soak, no amount of airflow will increase the TDP of a heatsink. The mass of the metal hasn't changed.

Liquid nitrogen AIO's don't exist. Anything more extravagant than an AIO water cooler is DIY territory, from custom water cooling to phase change cooling, dry ice, and liquid nitrogen. Let's not forget all the extensive anti-condensation prep work that has to be done in order to run any cooling methods more powerful than plain old water cooling. I'm not sure if this statement was meant as a joke or not, so for now, I'll treat it as a serious comment.

 

[The Original Ralph]

well, the consensus seems to be the mobo sagging or warping is not an issue

as to the D15, we're starting to get into serious tonnage there - noctua specs the wgt at 1320 grams or 2.910 lbs
the U14S is 935 grams or just over 2 lbs. The D15S comes in, surprisingly at 1150 grams or 2.5x lbs. But both D15s would be a pain to clean properly.

the U14S can also be boosted some with a 2nd fan, and if temps are still an issue, i'll look around for a nitrogen AIO system - someone has to be offering one, somewhere.

tks for the response

I've heard of people propping up their sagging D14's and Prolimatech Mega-WTF-behemoths with fishing line, tying it to the top of the case as a support, or fabbing a support rest directly into the side panel directly opposite the heatsink tower. Heat sinks are far easier to clean than the higher fin density radiators used for water cooling. At least you can use a leaf blower and not risk damaging anything.

Strapping a second fan for a push-pull config on a single stack heatsink like the U14S (and U12S and Hyper 212 and all of these) isn't as effective as you'd think. It'll help a bit, but the extra weight, noise, power consumption, and upfront cost don't justify the 2-3C difference, especially with the cost of a Noctua fan. The dual tower heat sinks definitely benefit from a second fan though. Heat soak is heat soak, no amount of airflow will increase the TDP of a heatsink. The mass of the metal hasn't changed.

Liquid nitrogen AIO's don't exist. Anything more extravagant than an AIO water cooler is DIY territory, from custom water cooling to phase change cooling, dry ice, and liquid nitrogen. Let's not forget all the extensive anti-condensation prep work that has to be done in order to run any cooling methods more powerful than plain old water cooling. I'm not sure if this statement was meant as a joke or not, so for now, I'll treat it as a serious comment.

actually it was said half-heartly - and the thought had occurred to me about fabbing a support from carbon fiber - fishing line, even metal wire will stretch from prolonged exposure to weight, but it would support some. Carbon fiber is 10X stronger than it's equivalent in 4130 chromemoly - (used in aviation frame tubing and in gun barrels). But it'd be a hoot to lay up a thin strap nearly the width of the top of the radiator area, let it form over the top end of the heat pipes, like socks or condoms and transit into a strap going to the top of the case. Would pull some heat out of the heatpipes as well as support the cooler. I'll see what happens when i mount the cooler, if i see any movement on the cooler when the mobo is in the vertical position

tks for the info on the condensation aspect - hadn't thought about that. Considering nitrogen's liquid temp is somewhere down in the -300F range, i'm surprised someone hasn't come up with an AIO system, kind of a crude air conditioning heat exchange system but i don't know enough about gases, so there's probably some obstacles i'm not aware of.

in case anyone gets thinking about fabricating something from carbon fiber for their computer, keep one thing in mind - carbon fiber will conduct electricty, same as if it were copper - we had also made some CF briefcases about 15 yrs back - an IBM computer tech bought one to use for his tool case. Got a call from him that the case had slipped down into the computer cabinet and shorted out the motherboard. so FYI

 

[mcnumpty23]

yeah anything extreme like liquid nitrogen,helium etc actually gets too cold and condensation is a major problem

same with peltiers the cold side gets damn cold so again condensation would be an issue as well as having to dump the heat off the hot side

know nothing about carbon fibre reinforced polymers other than they are light and strong

but guess there must be a valid reason why no one has made a heatsink from them

or if there isnt a valid reason maybe you could start the next big thing by making one

 

[The Original Ralph]

i'm involved in too many projects now, and am trying to retire - looking at small RVs so the wife and i can travel and see this country

but to give you another example - back in 2005 i had a VW R32 with custom turbo (414 HP, 381 fl lbs torque). Oil pan was cast alum with a smooth flat bottom surface. I fabricated a CF sheet to the shape of the oil pan's bottom & sides - only went up the sides maybe 3", then fabricated a bunch of 1/2" deep 1/2" wide "C" channels from CF, cut them to the length of the oil pan, and bonded them to the sheet to create ribs - at the drag strip oil temps dropped 15F at the end of a run.

CF literally sucks the heat out of something and spits out into the ambient air