Reverse gpu fans

[Sambi7]

Has anyone tried this, is there any data on cooling performance? 

 

I have a strange setup where I have two graphics cards, one water cooled, one air. 

The air card sits below the water card so it has access to airflow. 

Looking at buying a new case with upside down mobo orientation, which would have the air card sit reasonably close to top vents for the case. 

I'd want hot air to exhaust from the top if possible, so reversing the graphics cards fans seems to make sense, but I don't know the effect that the loss of static pressure on the heatsink would have.

Has anybody got any, even vaguely relevant, experience? 

[mxk]

You can't do that. The fans are made to spin the one way.

[Sambi7]

40 minutes ago, mxk. said:

You can't do that. The fans are made to spin the one way.

I can...

I take the fan off, I turn the fan over, I put it back. 

[aezakmi]

2 minutes ago, Sambi7 said:

I can...

I take the fan off, I turn the fan over, I put it back. 

it won't fit

 

I tried

[Addvilz]

It is unlikely to give you any measurable benefit, and in fact, could have negative effect - air is generally pushed on to and through heat sinks as it generates more turbulence and air pressure, and thus, improves heat transfer. Pulling the air through the heat sink has the opposite effect entirely, and given that the air will travel path of least resistance, it will bypass parts of the heat sink entirely. It really depends on the heat sink design however, but this applies as a general rule.

 

Quote

You can't do that. The fans are made to spin the one way.

That's what we have zip ties for. 

[Sambi7]

2 minutes ago, aezakmi said:

it won't fit

 

I tried

That's what a dremel is for. 

[Sambi7]

2 minutes ago, Addvilz said:

It is unlikely to give you any measurable benefit, and in fact, could have negative effect - air is generally pushed on to and through heat sinks as it generates more turbulence and air pressure, and thus, improves thermal transfer. Pulling the air through the heatsink has the opposite effect entirely, and given that the air will travel path of least resistance, it will bypass parts of the heat sink entirely. It really depends on the heat sink design however, but this applies as a general rule.

 

That's what we have zip ties for. 

I understand the loss of static pressure on the heatsink is going to raise the temps, I'm just curious as to how badly. 

If I keep the card stock, it's going to stop the exhaust from the top of the case and blow hot air all over my water cooled card, which is not a full block, and really piss off its vrms. 

[Addvilz]

Just now, Sambi7 said:

I understand the loss of static pressure on the heatsink is going to raise the temps, I'm just curious as to how badly. 

If I keep the card stock, it's going to stop the exhaust from the top of the case and blow hot air all over my water cooled card, which is not a full block, and really piss off its vrms. 

Wild guess? I would expect 10-15% increase or more.

 

And I was only partially joking about those zip ties. Set up some kind of test in a cardboard box and test the airflow and temps. That's the only (realistic) way to tell.

[WoodenMarker]

It depends on the system airflow and graphics card cooler. If the graphics card fans are close to the exhaust at the top, it could be as good or better than dumping heat into the case that isn't quickly exhausted. It could also be worse and it's hard to tell without trying it out for yourself.

[Sambi7]

Just now, Addvilz said:

Wild guess? I would expect 10-15% increase or more.

 

And I was only partially joking about those zip ties. Set up some kind of test in a cardboard box and test the airflow and temps. That's the only (realistic) way to tell.

Mounting the fans upside down isn't too much of an issue, I should be able to just flip them and screw them to the shroud. 

Might just have to try it out, the cardboard case idea is possible, but super impractical due to the water loop. Would mean draining, dissambling the loop, leak testing again, not a small job. 

[Sambi7]

6 minutes ago, WoodenMarker said:

It depends on the system airflow and graphics card cooler. If the graphics card fans are close to the exhaust at the top, it could be as good or better than dumping heat into the case that isn't quickly exhausted. It could also be worse and it's hard to tell without trying it out for yourself.

Thanks for the advice, it seems like the only way to find out is to do it and see. 

The practical use case for reversing gpu fans is so small I just can't find any usual info. 

There's some discussion, but I can't find any actual numbers. 

If/when I end up doing it, I'll chuck the outcome on here incase anyone else finds it relevant. 

[LogicalDrm]

12 hours ago, Sambi7 said:

Thanks for the advice, it seems like the only way to find out is to do it and see. 

The practical use case for reversing gpu fans is so small I just can't find any usual info. 

There's some discussion, but I can't find any actual numbers. 

If/when I end up doing it, I'll chuck the outcome on here incase anyone else finds it relevant. 

You could do some testing with top-flow CPU coolers. The principle is same, air is pushed through heatsink and rest is flown over VRMs as passive cooling. Reverse fan on that and you can see how much effect might be.