View Single Post
  #3 (permalink)  
Old December 1, 2012, 07:03 AM
Usersd Usersd is offline
Join Date: Nov 2012
Posts: 2

Originally Posted by Ardric View Post
Using a big power rheostat (potentiometer wired as a variable resistor) is really the wrong way to do it. It's inefficient; it throws off a lot of heat into the case, which is counter-productive. It increases the power supply impedance to the fan motors, which they can cope with but don't really want.

Most fan controllers supply 12VDC but use pulse-width modulation (PWM), where they switch the 12V power on and off really rapidly, and the percent time in the ON state (called the duty cycle) is varied to control the fan speed. This is much more power-efficient than a rheostat and makes less heat. It's still not a very nice thing to do to the fans though. Fans are motors, and motors are inductors, so in theory the motor should be ok smoothing those power spikes out again. But in practice, fan motors are trying to be more clever than that, and it's hard to be clever when someone keeps turning your power supply off. That's why we're moving to 4-wire fan connections with a separate PWM wire... by seperating the control signal from the power signal, the fan motor can have more complex electronics and do more neato things.

Generally the power limit to a rheostat is thermal dissipation, aka heat output, and it's measured in Watts. Work out how much current max you'll connect to the rheostat, then use watts=volts*amps to find out how much wattage that is. A 3W unit can handle 3/12=0.25 A (250 mA) of current at 12V, assuming all the power ends up in the rheostat and none in the fan. That's a good assumption to be sure you always have some safety headroom. But 250 mA isn't all that much. A 3W rheostat is way too small for a chain of 6 big fans.
Wow, a lot more detailed than I was expecting! And It actually makes sense too, that's why most fan controllers don't mindlessly heat up with 6 or 7 channels, a rheostat being basically an adjustable resistance, has to dissipate heat, PWM works differently.

PWM seems the way to go for now then, and AFAIK, it's a simple circuit and there are a lot of guides specifically for PC fans, I might take a jab at it, just by looking at the schematic the electronics store downtown should have all the components I need. Making one to control my biggest fans in the case (1x140mm, 2x120mm) should be fairly simple, and should end up much better than using the big old rheostat.
The LED driver i was thinking about doing would involve a simple PWM circuit with a 555 timer I found in various places on the interwebs too, to act as a dimmmer.

Well, thanks for all the info, learned a couple of useful things today thanks to you!

Last edited by Usersd; December 1, 2012 at 07:13 AM.
Reply With Quote