It's a project. I get in to doing projects like this to show it can be done.
The card has a thermal chip that's not being used to full potential. If the hack on the chip is successful then, and if I'm lucky, the true RPMs of the fan will not only be lower due to the new fan being 120mm and not having to go as fast as the stock fan, and if real lucky the RPM's will be read out in HWM. The smaller stock fan is loud and I want to get the noise level down.
The way they control the speed of the stock 2 wire fan is by turning on and off the power to the fan. The time between on and off and the duration of the on and off cycles is governed by the thermally controlled sensor chip. It has a PWM signal output and in this case that signal drives an interface transistor. All the transistor does is act like a switch to control the power to the stock fan. It's inefficient and noisy. When the power is off the magnets acting on the metal in the armature and the drag of the fan blades and the drag on the shaft in the bearings, slow the fan platter down rapidly. Then the next ON pulse comes along and you have to overcome inertia.
Like I said, it's inefficient, noisy, and puts unwarranted stress on the fan from the surge of the ON cycle.
With a 4 wire fan having it's own PWM input to its microprocessor chip, it uses that to decide which bank of coils in the armature gets energized and in what polarity. The power fed to the 4 wire fan is not switched so it remains constraint. Same voltage level regardless of the fans RPM's.
The current level will vary some but not really enough to be overly concerned about. My Cougar Vortex 120mm only draw 250 milliamps at max RPM's. That's about average for fans in that size class.
So in a 4 wire fan the coils are energized constantly. It's the current level and polarity of the current that is altered to make it go up or down in RPM's. A 2 wire just turns the power feed on and off to regulate the fan speed.
The card has a thermal chip that's not being used to full potential. If the hack on the chip is successful then, and if I'm lucky, the true RPMs of the fan will not only be lower due to the new fan being 120mm and not having to go as fast as the stock fan, and if real lucky the RPM's will be read out in HWM. The smaller stock fan is loud and I want to get the noise level down.
The way they control the speed of the stock 2 wire fan is by turning on and off the power to the fan. The time between on and off and the duration of the on and off cycles is governed by the thermally controlled sensor chip. It has a PWM signal output and in this case that signal drives an interface transistor. All the transistor does is act like a switch to control the power to the stock fan. It's inefficient and noisy. When the power is off the magnets acting on the metal in the armature and the drag of the fan blades and the drag on the shaft in the bearings, slow the fan platter down rapidly. Then the next ON pulse comes along and you have to overcome inertia.
Like I said, it's inefficient, noisy, and puts unwarranted stress on the fan from the surge of the ON cycle.
With a 4 wire fan having it's own PWM input to its microprocessor chip, it uses that to decide which bank of coils in the armature gets energized and in what polarity. The power fed to the 4 wire fan is not switched so it remains constraint. Same voltage level regardless of the fans RPM's.
The current level will vary some but not really enough to be overly concerned about. My Cougar Vortex 120mm only draw 250 milliamps at max RPM's. That's about average for fans in that size class.
So in a 4 wire fan the coils are energized constantly. It's the current level and polarity of the current that is altered to make it go up or down in RPM's. A 2 wire just turns the power feed on and off to regulate the fan speed.