Corsair SFX 450 Fan not spinning

[ShadowChaser]

I guess this is more of a peace of mind thing since the temps aren't ridiculous, but still really hot.

I noticed that the PSU in my Node 202 was still burning hot after I tore it down to replace the mobo and CPU (yay black friday) and the back of the PSU was hot enough that I couldn't hold my finger on it for more than 10-15 seconds. That was with a system with a power draw under load of about 250W. Now it's up to about 300W and I'm very concerned that the fan is not spinning while under load (folding, gaming, VR, etc). I can physically feel the heat radiating from the top of the PSU. Is this intended or did the fan die on me? Is there a way to tell?

 

Also, would it be worthwhile to upgrade to a SF600 (gold or platinum, doesn't really matter) with my current system?

8700K + H75

32GB DDR4-2666

Asus Strix Z390i

MSI GTX 1070Ti Aero

Samsung 970 Evo + 850/860Evo

2 Noctua P-12 fans

~300W from the wall (as measured by my UPS)

[tarfeef101]

If you're OCing hard (assuming you are cooling those parts well in a presumably small case), it might be worth a better PSU. Stock and basic OCing should be fine, though. 

As for the PSU fan, the units themselves can get warm. With my RM1000x in my testbench, even with a 4770k delidded and OCed on water + an rx580 OCed running synthetic load, it doesn't spin up. It feels hot, but not so much it'd burn some paper on top of it.

I would say look at some reviews (ie JonnyGuru) for your PSU. Figure out around where the fan should spin. Set your PC with a load equal or higher than that in terms of power draw. If the fan still doesn't spin, then perhaps you should be looking at an RMA. (ofc, you could save effort my contacting corsair, asking them when it should spin up, and measure based on that. although if it is based on temps, that could require equipment most people don't have. but common sense should still prevail. running those parts you have at max on an OC should totally result in a spinning fan unless your ambient temps are super low)

[ShadowChaser]

10 minutes ago, tarfeef101 said:

If you're OCing hard (assuming you are cooling those parts well in a presumably small case), it might be worth a better PSU. Stock and basic OCing should be fine, though. 

As for the PSU fan, the units themselves can get warm. With my RM1000x in my testbench, even with a 4770k delidded and OCed on water + an rx580 OCed running synthetic load, it doesn't spin up. It feels hot, but not so much it'd burn some paper on top of it.

I would say look at some reviews (ie JonnyGuru) for your PSU. Figure out around where the fan should spin. Set your PC with a load equal or higher than that in terms of power draw. If the fan still doesn't spin, then perhaps you should be looking at an RMA. (ofc, you could save effort my contacting corsair, asking them when it should spin up, and measure based on that. although if it is based on temps, that could require equipment most people don't have. but common sense should still prevail. running those parts you have at max on an OC should totally result in a spinning fan unless your ambient temps are super low)

So I did something that no one should ever do.

I powered 2 system with a single PSU.

I loaded the psu down with about 550W from the wall and it still doesn't spin up UNTIL it reaches 60C at the exhaust so who knows how hot the inside is getting

I guess that's good that it'll still spin up but only under the extremes that I don't subject it to.

Sometimes I wish I had a thermal imaging camera and not a laser thermometer so I can figure out where the hot spots are.

[tarfeef101]

Just now, ShadowChaser said:

So I did something that no one should ever do.

I powered 2 system with a single PSU.

I loaded the psu down with about 550W from the wall and it still doesn't spin up UNTIL it reaches 60C at the exhaust so who knows how hot the inside is getting

I guess that's good that it'll still spin up but only under the extremes that I don't subject it to.

Sometimes I wish I had a thermal imaging camera and not a laser thermometer so I can figure out where the hot spots are.

That's totally fine then. Most caps, FETs, etc are rated at AT LEAST 85C. Many at 105+ (caps). As long as the fan will spin up eventually, then you're all good. Corsair (and their OEM who I don't remember off the top of my head) chose that fan curve for a reason. You're all good

[Guest]

13 minutes ago, ShadowChaser said:

So I did something that no one should ever do.

I powered 2 system with a single PSU.

I loaded the psu down with about 550W from the wall and it still doesn't spin up UNTIL it reaches 60C at the exhaust so who knows how hot the inside is getting

I guess that's good that it'll still spin up but only under the extremes that I don't subject it to.

Sometimes I wish I had a thermal imaging camera and not a laser thermometer so I can figure out where the hot spots are.

You guys complain that the fan spins up too soon.

 

The complain that it doesn't spin up soon enough.

 

Guess what they say is true.  You can't make everyone happy.

[ShadowChaser]

1 minute ago, jonnyGURU said:

You guys complain that the fan spins up too soon.

 

The complain that it doesn't spin up soon enough.

 

Guess what they say is true.  You can't make everyone happy.

I dunno, I've always heard sub 70C is optimal for longevity yet the fan doesn't start pushing heat until the external temp is ~60C (internal temp 75+? Just a guess if it burns me)

Is this wrong or just old information?

[Guest]

17 minutes ago, ShadowChaser said:

I dunno, I've always heard sub 70C is optimal for longevity yet the fan doesn't start pushing heat until the external temp is ~60C (internal temp 75+? Just a guess if it burns me)

Is this wrong or just old information?

PSU doesn't care what the temperature is in your PC.  It only cares what the temperature is inside the PSU.

 

[Stefan Payne]

12 hours ago, jonnyGURU said:

You guys complain that the fan spins up too soon.

The complain that it doesn't spin up soon enough.

Guess what they say is true.  You can't make everyone happy.

That's why switch to switch off Semi Fanless is so great
Because with that, you can kinda.

[Guest]

9 hours ago, Stefan Payne said:

That's why switch to switch off Semi Fanless is so great
Because with that, you can kinda.

That's why adding cost to your product is so shitty.  Because then it's too expensive and people don't buy it at all.

 

[KuJoe]

22 hours ago, jonnyGURU said:

You guys complain that the fan spins up too soon.

 

The complain that it doesn't spin up soon enough.

 

Guess what they say is true.  You can't make everyone happy.

In the case of a PSU that's paired with an ITX case the fan curve should be spinning up sooner since it's 1/5th of the total number of fans in the case.

[Guest]

1 hour ago, KuJoe said:

In the case of a PSU that's paired with an ITX case the fan curve should be spinning up sooner since it's 1/5th of the total number of fans in the case.

In any case, if the overall cooling is dependent on the PSU fan, then you're doing it wrong.   If you ARE doing a build that is dependent on the PSU fan being part of the cooling solution, you don't buy a PSU with a ZeroRPM fan mode.

[Stefan Payne]

20 hours ago, jonnyGURU said:

That's why adding cost to your product is so shitty.  Because then it's too expensive and people don't buy it at all.

or just don't implement Semi Fanless (without a switch)

Because there are situations where you need or want the active cooling...

The dumb version with 2 pin, voltage regulated fan would "only" require two diodes and the switch and a line from either the 3,3V or 5V rail - et voila, a switchable Semi Fanless solution.

I did build something like that for my CWT SFX unit. with an old Schottky Diode from some older PSU (and used the 3,3V rail)...

20 hours ago, KuJoe said:

In the case of a PSU that's paired with an ITX case the fan curve should be spinning up sooner since it's 1/5th of the total number of fans in the case.

No, not true.

Not "an" ITX case like Node 202, Silverstone RVZ Series (all from 01-03), some Milos and some more, but your case.

 

So if you talk about stuff, pls mention what the heck you're talking about...
 

[KuJoe]

50 minutes ago, Stefan Payne said:

No, not true.

Not "an" ITX case like Node 202, Silverstone RVZ Series (all from 01-03), some Milos and some more, but your case.

 

So if you talk about stuff, pls mention what the heck you're talking about...

I'm sorry, it's probably the lack of sleep but you've confused me. To my knowledge this thread is about the Node 202 which both I and the OP are using.

[Stefan Payne]

20 hours ago, KuJoe said:

I'm sorry, it's probably the lack of sleep but you've confused me. To my knowledge this thread is about the Node 202 which both I and the OP are using.

Yes and in the Node 202 the PSU might sit in a different orientation but is not part of the airflow thing.

Like it is the case on the Silverstone Ravens as well...

[KuJoe]

2 hours ago, Stefan Payne said:

Yes and in the Node 202 the PSU might sit in a different orientation but is not part of the airflow thing.

Like it is the case on the Silverstone Ravens as well...

I thought every fan in a case is "part of the airflow thing", are you saying it has no impact to airflow in the case?

[Stefan Payne]

1 minute ago, KuJoe said:

I thought every fan in a case is "part of the airflow thing",

No, only if it influences that in a way.

That is not the case with more modern ATX cases where the PSU is on the bottom and also with many other cases.

With ITX it depends on the design of the case, where the PSU is. If it is in a seperate compartment then its not relevant, 

 

1 minute ago, KuJoe said:

are you saying it has no impact to airflow in the case?

Correct as the Fan is directly taking air from the outside and blowing it out again without ever interacting with the things inside.

[KuJoe]

9 minutes ago, Stefan Payne said:

No, only if it influences that in a way.

That is not the case with more modern ATX cases where the PSU is on the bottom and also with many other cases.

With ITX it depends on the design of the case, where the PSU is. If it is in a seperate compartment then its not relevant, 

 

Correct as the Fan is directly taking air from the outside and blowing it out again without ever interacting with the things inside.

We'll have to agree to disagree, I find that really hard to believe because the temps I am feeling from the PSU exhaust differ depending on the temps of the other sensors in the PC.

[OrionFOTL]

On 12/3/2018 at 4:38 PM, KuJoe said:

In the case of a PSU that's paired with an ITX case the fan curve should be spinning up sooner since it's 1/5th of the total number of fans in the case.

6 hours ago, KuJoe said:

I thought every fan in a case is "part of the airflow thing", are you saying it has no impact to airflow in the case?

It has no impact to airflow in the case, because the PSU fan isn't "in" the case. From what I can see on the photos of Node 202 online, you put the power supply in this spot, fan-down:

So the PSU intakes air through the ventilation mesh, then exhausts it through the back like the arrow shows. The air the fan sucks in and pushes out never reaches anywhere inside the case.

 

[ShadowChaser]

6 hours ago, KuJoe said:

We'll have to agree to disagree, I find that really hard to believe because the temps I am feeling from the PSU exhaust differ depending on the temps of the other sensors in the PC.

The temps of the components do matter. If the gpu or cpu is hotter it's probably running harder -> more power needed -> PSU delivers more power -> PSU gets hotter.

Also since the psu is metal, heat will transfer, though that's really negligable in this case (see what I did there hahahahaha kill me)