SFF System. PSU Fan Dying. Leave it?

[multiplexed]

I’ve never worked much with SFF systems. I have a Dell Optiplex GX270 sitting here. It’s an 8 year old, odd form factor PC. The processor fan is dead, likely for quite some time. The power supply is obviously not a standard ATX PSU. It’s a smaller height, very long PSU. Here’s a picture …
http://www.txcesssurplus.com/servlet/the-7906/NEW-Dell-OptiPlex-GX270/Detail

My problem is there’s a fan on the end of the PSU (the black part) that’s making a racket like it’s going to die soon.

My question is, is there any harm in just letting this fan die? I don’t think it’s moving any significant air. I can’t feel any air movement. There are no air vents in the case beneath the fan. Is it cooling something inside the PSU? It’s not doing anything for the case temperature.

Again, should I just let both of these fans die. The CPU fan has been dead for a while & it’s still running fine. I don’t want to sink $100 in this if it’s not needed. I’m thinking to just blow it out with some canned air and leave it be.


Edit: If I do leave this fan alone, and the PSU blows up or something, does this carry a risk of blowing up the motherboard as well? I'd rather spend the cash and not have this happen, but if it's not a significant risk then I may just leave everything be.

 

[Raiddinn]

If you let the PSU fan completely die, there is definitely going to be a risk the PSU will blow up and that it will take the motherboard or anything else attached with it. That means things like hard drives are fair game since they plug directly into the PSU.

For that matter, I would be very wary of letting my CPU fan die as well.

The CPU has special circuitry that causes the chip to handicap its own performance if it notices that its temperatures are too high, but if that failsafe ever doesn't work then your processor could straight up melt, potentially destroying the motherboard at the same time.

It may sound like I am being dramatic, but its far too common around here that people experience hardware damage that traces straight back to excessive heat, particularly inside the PSU.

If you want to avoid this expense, then make sure all your data is backed up and start saving for your next PC. I would highly doubt you have even as long as 6 months from the time your PSU fan dies, maybe not even 6 days. - Edit - Maybe not even 6 minutes.

 

[multiplexed]

I’ll error on the side of caution and replace the parts, but I’m not convinced that they are realistically needed.

Thanks for the reply though.

 

[13thmonkey]

if you weren't worried then you wouldn't have asked the question.

If you are not convinced then don't do it, come back in a few days/weeks/months and ask which bits are now broken.

If you are an electrical engineer you can probably just replace both fans for $10.

 

[Raiddinn]

For the processor you could do it pretty easily, however, for the PSU things would be a little more complicated.

PSU components all deteriorate over time. The farther away from 50% load the PSU is when it is on the faster this process happens. Especially so when the load approaches 100%.

If you just replace the fan, you would potentially clear up the most annoying symptom of the dying PSU without doing anything to address the erosion of the other internals.

PSUs are only built to last about 5 years for the most part, especially so for top mounted PSUs and SFF PSUs. I am guessing your GX 270 PSU is pretty much at that mark.

If you were an electrical engineer, I would suggest you use your knowledge to replace the coils and stuff too.

If you aren't convinced that an old PSU can damage computer hardware, I am sure there are hundreds of examples here on Toms Hardware where just that has happened, and that is in the last year not since the board was created. Feel free to do some searches. Start with "BANG", "Sparks", or "old computer won't turn on" as your search criteria.

 

[multiplexed]

I would completely agree with you guys if this was a standard tower / ATX power supply. It would definitely need replaced in this situation.

However, this is a small form factor power supply. The fan is in the very corner of the power supply, so it’s impact may be minimal. There are no vents in the case beneath the fan, so it’s not cooling the case at all. I feel zero air flow on top of the fan. Plus, I think that when I close the lid, there’s a 3.5” hard drive that comes down directly on top of the fan. That has to minimize any effect of the fan itself.

My main concern is, is this fan ‘realistically’ doing anything? My techie side says yes, but my practical side says it’s a waste.

I’ll error on the side of caution and replace it, especially since it’s not expensive to fix. But again, to me this fan doesn’t appear to be doing anything useful.

Anyway, I’ll be ordering parts soon. Thanks for everyone’s replies. I do appreciate them.

 

[Raiddinn]

The PSU fan shouldn't be taking hot air from inside the case into the PSU in order to cool the motherboard and other components.

It should be taking in cold air into the PSU in order to cool the PSU components.

 

[13thmonkey]

The PSU fan shouldn't be taking hot air from inside the case into the PSU in order to cool the motherboard and other components.

It should be taking in cold air into the PSU in order to cool the PSU components.

no, a normal psu takes air from the case through the psu to the outside.

some cases now allow you to have the input side of the psu through a vent in the base of the case, the output is still to the outside.

 

[DJDeCiBeL]

no, a normal psu takes air from the case through the psu to the outside.

some cases now allow you to have the input side of the psu through a vent in the base of the case, the output is still to the outside.

+1 A top mounted PSU (in a normal older style case, like mine) doesn't allow for bringing in fresh air and it vents outside the case, regardless.

 

[Raiddinn]

"Normal" cases are designed with an extremely flawed methodology.

PSU fans are intended to take in cold air, run it across PSU components to cool them, and then force that air out the other side of the PSU.

Early ATX specifications wanted to use these fans to do double duty and cool the other computer components at the same time.

That doesn't mean it is or was ever a good idea.

 

[DJDeCiBeL]

"Normal" cases are designed with an extremely flawed methodology.

PSU fans are intended to take in cold air, run it across PSU components to cool them, and then force that air out the other side of the PSU.

Early ATX specifications wanted to use these fans to do double duty and cool the other computer components at the same time.

That doesn't mean it is or was ever a good idea.

Not arguing that point at all. It IS stupid.

 

[13thmonkey]

"Normal" cases are designed with an extremely flawed methodology.

PSU fans are intended to take in cold air, run it across PSU components to cool them, and then force that air out the other side of the PSU.

Early ATX specifications wanted to use these fans to do double duty and cool the other computer components at the same time.

That doesn't mean it is or was ever a good idea.

NO but it does mean that is exactly what happens, and why they are rated to work at 40C or 50C for the better ones to take account of how they were designed.

 

[Raiddinn]

Better PSUs don't work at 50c because the maker expects them to be top mounted and used to cool PC components. They are made so because better quality parts allow the whole PSU to work better and longer regardless of what situation it is in.

It is still a bad idea to use the PSU to cool the PC regardless what the PSU is rated at. Anyone who wants their computer to be stable and long lasting wouldn't want to rely on the PSU for cooling PC internals.

The last thing we need to be doing is to encourage people to use them in a way other than the manufacturer intended them to be used, whether it is common to use them wrong or not.

 

[13thmonkey]

that is what they were built for at the start, they were the only extraction fan in the case, there never used to be fan locations, the better PSU's are rated to 50C and will perform at 50C. Std PSU's are rated at 40C and should be loaded tested at this temp, few few sites load test at rated temp.

What we should not be encouraging people to do is running a case with no ventilation, push and pull.

 

[Raiddinn]

Short sighted decisions by the people who wrote the ATX specifications aside, what we should be doing is helping people to build PCs that work both on day 1 and on day 1500.

If nobody here will try to argue that thing B is better than thing A, then there is no reason to encourage the doing of thing B.

The only valid reason to do anything different is if some sort of restriction is in place that doesn't allow for thing A.

I steer people away from doing the wrong thing on a daily basis and one of those wrong things is using PSU fans for airflow management. That idea was fail from the minute PCs started generating heat and is certainly still so in 2012.

 

[13thmonkey]

I don't disagree with the need for better ventilation, but the psu is built and designed to work on internal case air. in 95% of cases you have no choice.