It works only when needed by the system.
Turn on hot end heating and bed heating and you should be able to hear it turning on
Shape of the wings do suggest that it is meant to push the air in the chassis.
Mine blows air out but didn't take it apart to check blade position
You can turn it whatever way you want and see which way it cools better
Gt2560 rev B noise Fan
Re: Gt2560 rev B noise Fan
My plans to reduce noise is the use an adjustable voltage regulator (sub $1 DC-DC module) of which I have plenty. The cheapest ones are dangerous for some of my other purposes as they will let input voltage thru when they die. But for this it is no problem, just means full fan speed. I have that inside a couple server PSUs to slow fan down a lot.
The other thing that can be done when slowing fan down is to guide airflow over the board, so I do consider making cardboard/printed boarder around the GT2560, with cutouts where the drivers are. That would force all airflow over them, requiring less. Fan ducts are a standard on brand name servers. Basicly that would be a strip alongside the edge facing the big space in the case. Now the fan pulls air out with a flow that is pretty low over the drivers. So more cooling should be obtainable by lowering speed, and have all the airflow actually cool the drivers.
Should be possible to add a thermistor in the circuit as well, fan runs slow low temp, and faster at high temp. Let us assume the fan uses 0.1A @ 12V, that is an internal resistance of 120Ohm, so we likely need something that have a max resistance of say 480 Ohm -> 1k Ohm. (1/4-1/8 speed minimum). Then put the thermistor near a board we know gets hot, like the extruder, x or y board.
The other thing that can be done when slowing fan down is to guide airflow over the board, so I do consider making cardboard/printed boarder around the GT2560, with cutouts where the drivers are. That would force all airflow over them, requiring less. Fan ducts are a standard on brand name servers. Basicly that would be a strip alongside the edge facing the big space in the case. Now the fan pulls air out with a flow that is pretty low over the drivers. So more cooling should be obtainable by lowering speed, and have all the airflow actually cool the drivers.
Should be possible to add a thermistor in the circuit as well, fan runs slow low temp, and faster at high temp. Let us assume the fan uses 0.1A @ 12V, that is an internal resistance of 120Ohm, so we likely need something that have a max resistance of say 480 Ohm -> 1k Ohm. (1/4-1/8 speed minimum). Then put the thermistor near a board we know gets hot, like the extruder, x or y board.