Which way should my case fan blow ?

There is a hole in my computer's case beneath the PSU (top rear). Which direction should the airflow be for best cooling ? [into the case/out of the case].

Further details below.

I have decided that my system might benefit from a case fan in the recent hot weather. At present my PC has two operating fans, one an exhaust fan in the PSU (what would be the rear top) and the other the CPU fan, which blows air from the space closest to the removable side panel through the CPU heatsink fins towards the CPU (pic of m/b here, PSU lies at the bottom right and exhaust direction is down). There are some small holes in the side panel, and a perforated honeycomb hole in the rear of the case below the PSU fan grill. These things entirely determine the airflow through the PC as it was purchased (with only PSU & CPU fans, no case fan).

Recently I noticed that around/in the honeycomb vent in the back, there were 4 small holes in an 82.5mm square pattern. I thought that these must be for the purpose of mounting a a case fan over the honeycombed hole. I have bought a 92mm Akasa Smartfan (how do you use the rubber pins to mount the fan ?) for the purpose.

Conventionally a rear fan is to exhaust air from the case, but there are no vents apparent through which 'cool' air can enter the case, except the few small ones in side of the case. Presently the only forced airflow is through the PSU out of the case, with air being drawn in/entering (I assume) the case though the honeycombed hole because of the minute pressure difference caused by this. Therefore it makes no sense to use a fan to extract air from the case though a hole which otherwise would be the main source of cooler air ?

Dumb_Question
30.July.2013

Compaq Presario S5160UK DT261A under XP/SP3
Processor - Celeron 2.7 GHz
Motherboard - MSI MS-6577 v2.1
RAM - 1GB + 512MB (1GB +1GB max) DDR PC2700
PSU - Octigen 300W model 10270PSOTG ('upgraded' from original Bestec 250W PSU [in 2011?])
Nvidia GeForce 6200 graphics card in AGP slot.

 

You have an 82mm hole so you bought a 92mm fan, hmm, it'll probably be hackable to fit - it should be set to exhaust the air - it's in a hotspot there.

There should be inlet vents around the front panel, they could be hard to locate though, also there's usually a space between the front panel and the front chassis to fit a 120-140mm fan. Set that front fan to blow inwards - you should get one for Summer use.

If the negative pressure is high enough (high CFM fan set to exhaust), there's not usually any problem with cooling air ingress - air rushes to the lowest pressure area - it will find the easiest way in naturally.

 

This thread might help you understand and expand on Satrow's advice.

http://www.sysopt.com/showthread.php?207271-PSU-at-its-final-stage

Hope it helps.:wave

 

Don't worry about over and under pressure (unless you just want to learn about it). It's too confusing for most, and not important to understand. They matter, but not near as much as "flow" through the case.

Generally, you want good "front-to-back" flow of cool air through the case. So fans in front need to draw cool air in, and fans in back need to exhaust heated air out. Since heat naturally rises, some cases support, or can be modified to support "blow hole" fans - fans mounted on the top of the case. They should help that heat escape by blowing the air out (up).

Unless your case supports the larger 92mm fan (already has the necessary screw holes) I would avoid using the fan and instead, get an 80mm. Otherwise, you will need to drill holes in your case - and that REQUIRES you remove EVERYTHING out of the case or risk metal filings falling on/in your electronics - a very bad thing.

Finally, since it is the case's responsibility to provide an ample flow of cool air, if your case doesn't, I recommend replacing the case with a good case, that does. A good case will provide many years of support through many upgrades - thus is a good investment - and along with a quality PSU, form a solid foundation for just about any computer.

I like Antec cases because they are not too flashy (so they don't draw attention to themselves, or go out of style), sturdy, true (exactly 90° bends), have lots of good fan support, and many have removable, washable air filters - a requirement for all my builds.

 

PCSO2007: I have read the posts in forum topic you linked to, and also looked at a (dissembly) picture of a dx6100. Apart from the thread raising all sorts of questions (questionable assertions ?), it certainly strongly implies that HP's thermal designers are useless at their jobs and don't know their a*** from their elbow (which may be the case, but I'd assume that these guys have much experience with this sort of problem and access to experimental techniques and sophisticated computer codes for modelling). The main purpose in the HP design (as I understand it !) is to get cool air straight from from the rear vent to the CPU heatsink (using the aid of a rear fan). I understand that some HP PC's and some Dells too, in fact have a hood giving a tunnel straight into the case from the case wall so as avoid the cooler air from outside the case being 'contaminated' with warmer air from inside the case. After a larger fraction has passed through the CPU heatsink, I suspect that most of it passes over the PSU components and out of the case.

It seems as if the airflow is exhausting through the front vents, but I for one, can't say this: it is likely that a (weaker, because it is not fan-assisted) current of air passes over the bottom of the lowest disk drive, along the rear of the other drives, and exits through the PSU. Anyway, without a front fan, whichever direction it is in is much smaller than the airflow through the rear holes. This has the advantage of the best CPU cooling. If the CPU heatsink is as in my PC, the air moves in a layer near the removable panel and parallel to it, before being directed in a perpendicular direction towards the CPU on the other side of the case, and blown out of the CPU heatsink towards the PSU and (perpendicular to) the PCI-E 16x slot (an AGP in mine, components and heatsink on my GeF6200 unfortunately on the other side facing bottom of case).

On the other hand, the advantages of the airflow pattern given by one of the responders (front-to-rear) in a useful illustration is that you appear to get better cooling of the PSU, and possibly of any cards you have in the PC (the PCI-E 16x might be an exception, depends on the temperature of the chips on that card vs temperature of air out of CPU heatsink), and certainly of the lowest drive. I really can't tell which is better because in either case most of the intake air goes through the CPU heatsink because of fan on top provides the easiest route. As one of the posters remarked, try both and see which lowers the temperature more, situating the computer where it would normally be located.

As Digerati says it doesn't much matter the case is under pressured or overpressured, the air isn't going to keep to keep on coming out or going in. You might argue that under pressured is best as air always 'wants' to get in, but it's still got to get out. (Overpressured, the air 'wants' to get out, thus drawing more cool air in)

I've looked my PC. I was incorrect about the number of vents, there are two mesh covered slots in the front panel behind which is a very holey section if the case (the lower two in the picture on this page. There are some smaller holes in the case there, possibly for mounting a fan ??? between the front panel and case, or inside the case if there's space in either place (I'm averse to doing either, it looks too fiddly). I've looked at the original and current PSU. The only holes I can see in the current one are a large number in the front of the PSU case (being opposite the fan end). In the original PSU, it has holes there, and in the rear half of bottom surface of the PSU case (facing downwards). There is only a small gap between the end of the PSU and the CD drive, less than 80mm, so it would be impractical to put a fan hole(s) there, and if it were an exhaust fan, this would deprive the PSU of air, especially with the present one. Thanks for stressing the importance of avoiding metal 'filings' in your PC !

A further note, alluded to earlier, is that in my PC the fan is configured to blow air through the heatsink from the "top" towards the CPU, and the fins are oriented to exhaust this air from the heatsink towards the AGP card and bottom of the PSU. With reference to the diagram on the webpage referenced by PCSO2007, the fins on the heatsink in my PC are thus oriented to block airflow through the heatsink in the airflow direction shown in thate diagram, forcing it to pass through the top and perpendicular sides of the h/s, thus avoiding either drawing the air up preferentially from the side the air is coming from, or blowing air in the opposite direction to the direction illustrated (an observation only).

One final note is that my PC is not the correct way up; I can get easier access to the inside with the side panel up (and to the numerous rear connectors with the rear facing me instead of the front facing me).

So it looks like I'll have to do an experiment to determine which direction the rear-mounted fan should direct air for best cooling - my thought is that it won't make a distinguishable difference

Thank you for your recommendation for Antec cases, Digerati

Dumb_Question
31.July.2013
Compaq Presario S5160UK DT261A under XP/SP3
Processor - Celeron 2.7 GHz
Motherboard - MSI MS-6577 v2.1
RAM - 1GB + 512MB (1GB +1GB max) DDR PC2700
PSU - Octigen 300W model 10270PSOTG ('upgraded' from original Bestec 250W PSU [in 2011?])
Nvidia GeForce 6200 graphics card in AGP slot.

 

Exactly - and take the heat with it. That's why flow is more important. Having all the fans (front and back) push air in (over pressure) or draw air out (under) is not conducive to good "flow" (at least in my mind).

In fact, (again, in my mind) for the air to be pressurized (over or under) air movement would become somewhat stagnant - and that is hardly a good thing when it comes to extracting heat.

 

I had that very thought when I writing, but I thought I'd said enough :-D

Maybe that is why you should always try to maintain a pressure-neutral situation for best airflow, though unless things naturally reach that point, it's difficult to see how that could be arranged - if you have three fans in you case, that's three variables in a temperature optimisation, before you even what temperatures and their constraints you need to optimise !!

Dumb_Question
31.July.2013

 

LOL - well, I tend to say more than I need - but oh well.

The problem is cases are not sealed. Air is sucked in through any and every opening including through slits around the CD/DVD tray, and any port, any vent. So you cannot simply install one fan in front and one fan in back and declare equilibrium. Another reason to not worry about over and under stuff - just enough stuff.

 

An update & result report

I installed (well, to call it an installation is a bit of overgrandisement) a rear case fan (one of these) as an extraction fan. It substituted the fan I had been kind of using to cool my graphics card (only one power connector until I get an adapter for 4 pin large Molex to mini-molex (KK series) to power another fan direct from PSU).

As a result the CPU temperature has fallen by about 8C and the graphics card temperature has risen by about 5C

(temperatures determined by Speedfan, with which I can also control the fan speeds - contrary to some reports I find that the temperatures measured by this [ver 4.45] are in agreement with those reported by other programs)

Don't know if it afffected the PSU temperature.

Wonder what will happen in intake geometry ? An advantage of extraction is that there less need if any for a dust filter (which I should certainly put in if in intake mode).

Dumb_Question
9.August.2013
Compaq Presario S5160UK DT261A under XP/SP3
Processor - Celeron 2.7 GHz
Motherboard - MSI MS-6577 v2.1
RAM - 1GB + 512MB (1GB +1GB max) DDR PC2700
PSU - Octigen 300W model 10270PSOTG ('upgraded' from original Bestec 250W PSU [in 2011?])
Nvidia GeForce 6200 graphics card in AGP slot.

 

Note matching the temp to correct sensor name is a problem with many 3rd party monitoring programs. For example, it may "correctly" say 46°C, but the program may say it is for System rather than the CPU, for example, or maybe just Fan1. But Speedfan, as do the others, allow you to edit the sensor names to correct the issue - so a minor annoyance.

Ummm, no. That is not correct. The extracted air must come from somewhere, right? It does not originate in a tank of clean, filtered air inside the case. As I said in my last post, the air the rear fan exhausted was pulled into the case through all the cracks, crevices, vents and other openings, along with the dust floating around your room. There is no reason to expect a rear fan to suck in less dust than a front fan. It just comes in from all directions, rather than primarily the front fan vent.

CFM is CFM - whether pushed along or pulled along.

 

Good point Digerati. The inflow must equal outflow whatever. I failed to mention that the inflow when the fan is extracting would mostly be through the two 'slots' at the front which have a fine (?) grille over them - not a filter admittedly, but pores are very much smaller than in the honeycomb pattern on the hole in the hole in the rear through which air would be drawn in in intake configuration.

Dumb_Question
9.August.2013
Compaq Presario S5160UK DT261A under XP/SP3
Processor - Celeron 2.7 GHz
Motherboard - MSI MS-6577 v2.1
RAM - 1GB + 512MB (1GB +1GB max) DDR PC2700
PSU - Octigen 300W model 10270PSOTG ('upgraded' from original Bestec 250W PSU [in 2011?])
Nvidia GeForce 6200 graphics card in AGP slot.

 

Right - which means it will not take long for a layer of dust to build over those "pores" which will cause the fans to try to draw in cool air through every other opening, including your optical drives (which, btw, creates a dusty lens). USB ports, headphone jacks, even around front panel buttons will collect and trap dust.

Sadly, unless your computer sits in a HEPA filtered room, dust will always be a problem - and a greater problem for those with multiple people shedding dead skin in the house, users with pets, users with rug rats running about stirring up dust. If there are smokers in the house (or worse, around the computer), the problem is exacerbated because tar and nicotine is sticky. Cat dander is the worst, however, as the oils are even stickier than tar and nicotine. With no sticky components, a lot of dust simply travels through the case with the air flow. But sticky components in dust stick to vents, heatsinks, and anything else they touch - then reach out and grab on to any dust particle that comes floating by.

I have seen long neglected computers of cat owners so caked with layers of oily cat hair, dander and dust, I almost had to take a putty knife to scrape it out - and being allergic to cats - that is not a fun task.

Because "heat is the bane of all electronics" and because a layer of heat trapping dust can form so quickly, my general recommendation is for all computer users/owners to inspect case interiors monthly, and clean as necessary.

And for the record, a primary reason I do NOT like notebooks (or notebook makers, actually) is because notebooks are NOT designed for easy user access to the deep interiors for a proper cleaning - without major disassembly.