• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Flipping PSU fan

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

cyberkost

Member
Joined
Feb 25, 2010
Location
New York, USA
I have one of those setups in which the PSU is directly above the CPU socket. This severely limits the CPU cooling solution choices. What I'm thinking of is getting something like Thermalright AXP-140 and reversing PSU fan such that air is taken through PSU's externally exposed side and pushed onto the mobo through the AXP-140's fins.

Has somebody tried to reverse the airflow of the PSU? Are PSU fan power and RPM control wires soldered in or there's a connector (much like on mobos) these days? (I know they were almost universally soldered in the past) ... I guess I can potentially flip the fan w/o disconnecting the wires, but the next step is actually to get the fan controlled by the mobo and not by the PSU ...

Any advice/experience will be appreciated.
 
There really should be no problem with flipping the fan . As for changing the wires if you are not ok with resouldering the wires there are other options ( But splices, Wirenuts) . If you are unsure with going to the pads on the PCB just cut the wire and work of that (very easy).

When you are in there watch out for the Capacitors they can hold the charge for a long time and could be deadly
 
On the capacitors -- how long is "long"? I'd think that once the LED on the mobo (the one that goes off a few seconds after the power switch is flipped off) is off it would be safe ...
Also, will the PSU "feel weird" once the fan is "amputated"? The dilemma here is that I want to cool off both the PSU and the CPU with one fan and the fan can only be controlled from one place (either the PSU or mobo). Perhaps I can just get a quiet fan hook it up to 12V w/o any control?
 
Why not duct the cpu fan to the psu and let the psu fan pull air through the cpu cooler?
Don't void the warranty that way.

Capacitors can hold a charge for days. Hopefully PSUs are designed to bleed the charge out faster though.


EDIT:
That said, one 160v jolt from a PSU cap is highly unlikely to be fatal or even cause injury.
NOT fun, but the dangers are overstated.
Now the caps you find in CRTs and microwaves and such are a different story, that's 14,000 volts.
 
Last edited:
The case is Lian-Li PC-Q07. A picture with side off won't really help, as the view will be blocked by the PSU and you can't get the angle for a good shot. I have core i3-530 in there and the stock HSF is audible at very high loads ... I'm just thinking if there's a $50 (or less) solution to that.
 
On using a duct -- I thought of that, but the space between the mobo and the PSU is really inaccessible in PC-Q07 ... The AXP-140 almost flush with the PSU is in effect making (a very short) duct ... I just wanted to flip the fan so that (relatively) cool air blows toward the mobo, helping other components as well.
 
Absolutely Console, but to get that amp through the skin takes lots of volts, the skin has a lot of resistance.

Frequency is the only reason wall electricity is nearly as dangerous as it is, if it were DC it would take much longer to kill you.

That's why the 14kv caps are more dangerous, more volts to pump the amps through you with.
On the extreme end, i've gotten 70k thumb to thumb before. Not fun.



As a test, go grab both terminals of your car's battery. That's 12 volts, and the battery will happily put out 500 amps, maybe as much as 1000 amps. Absolutely nothing will happen, as 12v isn't enough to punch through skin.
 
Bobnova, you started by saying the right thing, but got it all wrong at the end. Let me try to make it clear so that we can get back on topic:
1. The biological effect is determined by the magnitude of the current and the time it acted (+ the actual path through the body -- toe-to-toe may leave you w/o kids, but thumb-to-thumb may actually make you heart stop)
2. Dry skin has a resistance of a couple of MOhms. It is this resistance that determines the magnitude of the current that would flow through, given the voltage applied (nothing too serious should happen if one grabs the terminals of 12V car battery, as the current should be tiny, say 12V/2MOhm = 6microA ... I don't know where you got 500 amp as that would actually be lethal with a good likelihood)

Back on topic -- did anybody open up OCZ ModStream 600W PSU?

Absolutely Console, but to get that amp through the skin takes lots of volts, the skin has a lot of resistance.

Frequency is the only reason wall electricity is nearly as dangerous as it is, if it were DC it would take much longer to kill you.

That's why the 14kv caps are more dangerous, more volts to pump the amps through you with.
On the extreme end, i've gotten 70k thumb to thumb before. Not fun.

As a test, go grab both terminals of your car's battery. That's 12 volts, and the battery will happily put out 500 amps, maybe as much as 1000 amps. Absolutely nothing will happen, as 12v isn't enough to punch through skin.
 
As a test, go grab both terminals of your car's battery. That's 12 volts, and the battery will happily put out 500 amps, maybe as much as 1000 amps. Absolutely nothing will happen, as 12v isn't enough to punch through skin.

you sure about that? I have had burns from contact across the battery. As in noticeable red marks which ended up scabbing over.
 
and as an electrician I would rather get hit with AC any day then DC (it grabs and doesn't let go )
 
you sure about that? I have had burns from contact across the battery. As in noticeable red marks which ended up scabbing over.

Absolutely positive, i touch 12v constantly as an auto mechanic.

cyber: The battery can put out 500+, but you don't die due to the resistance of your skin, as my post stated.



I doubt anybody has taken one of those PSUs apart.
 
Cyber as I posted before is should be pretty easy to flip the fan + put it on a controler .

@Bob I think you are thinking of cold cranking amps
 
@ cyber, I understand your idea and it would be as simple as switching the fan, but psu's put out a very large amount of heat when they are under load (any system load) and what you would see is a huge increase in temps any time the system was stressed. You could only be stressing the gpu and the cpu temp would rise horribly because the psu was loaded more and pumped hot air in. Hot air has much less capacitance for heat than cool air, obviously. I think that bobnova probably had the best idea. A duct around the cooler up to the psu would pull air through the sink and into the psu. Albeit, either option, one component gets the short stick and has higher temps.
 
I have i3-530, 4x2.5" HDDs and 1x slim DVD (no discrete graphics) + a couple of LEDs -- do you think this will come close to stressing 600W PSU? My thinking is that it won't, although I see PSU fan spinning all the time (perhaps it's just the darn heatwave in Northeast).

As for the duct ... again, I thought of that .. and came to the conclusion that it's not quite possible in the case used for this build (LL PC-Q07)... I should be able to get around to posting some pics on the weekend -- hopefully, you'll agree with me after seeing them that duct is more than just tough in there.
 
Not even close, your system probably uses less then 150w at full load.


Amps and ohms and stuff: That's exactly what i'm saying. It can if the resistance is low enough.
 
Actually, I have a UPS that shows power usage .. will check how much it uses when it calculates 4x md5sums and rips a DVD to my RAID10.

As for 500A -- I don't think you'd be posting if that much ever flowed through your body ;)
 
Back