Does a cheap fan controller reduce fan energy consumption?

[chazman113]

Just finished my budget build and went with a very nice fractal designs case I found on slickdeals.net. It comes with a 3-fan controller (Just a little pci sliver with a knob on the back). I picked up a 4 pack of coolermaster fans for $13. The specs say they draw about 1.8w each. I don't really think I need extra fans in the case but the slots are available. My question is, if I use the controller and turn the speed down, will it actually draw less power than just having them on 100%? 4 fans @ 1.8 watts 24 hours a day comes out to about .50cents a month according to some calculators I was using, but upwards of $6 a year! hah

Thanks
Charley

[zburns]

My question is, if I use the controller and turn the speed down, will it actually draw less power than just having them on 100%?

Answer is YES! Power (watts) = Voltage (volts) X Current (amps). Reduce either voltage or current, or reduce them both, and power is proportionally reduced.
------------------
If you do this run this test after doing so. Without using the new added fans, check your cpu temp in the BIOS under highest (or close to the highest) load (games if you do games). Then try the new fans in addition to present case fans and check cpu temp. If cpu temp is 'really helped more than 10 %' use the new fans; otherwise use the original setup. Any fan addition is going to add to case noise level.

[chazman113]

Answer is YES! Power (watts) = Voltage (volts) X Current (amps). Reduce either voltage or current, or reduce them both, and power is proportionally

But if the fan controller uses some kind of resistor wouldnt the excess power just be released as heat? Or do they change the voltage? This one has a smooth action that seems to turn "something" down in a smooth fashion ie no clicks for each setting

[zburns]

You are correct in that the resistor would heat up and waste energy. Other circuits, both AC or DC can be used to control the fans speed. Your control 'knob' via additional circuitry will 'smoothly' lower the voltage on the fan. The max voltage would probably be +12 volts DC. The voltage could also be an AC voltage equal to that required to run a fan with AC volts; however, probably the ATX standard might require (?) that any ATX form factor computer would have fans run on DC volts only.

If AC voltage, a circuit of 'back to back' SCRs (silicon controlled rectifiers) can 'selectively' reduce the amount of ac voltage to the fan and, therefore, reduce the speed. The AC voltage has the shape of a continuous 'sine wave' provided the SCRs are turned 'fully on' 100% of the time. However, each SCR must 'fire' each 1/2 cycle; an SCR can be triggered so that 'less than a half cycle' is supplied thereby, limiting the 'energy' to the fan and slowing it down. If the SCRs operate 'wide open', the fan gets full voltage and runs at max speed. The nature of an AC circuit just described may generate 'unwanted noise' (interference) and may not be suitable for computers; whereas DC circuits may avoid this problem altogether and therefore, be the choice of voltage type to be used.

If DC voltage, the control knob can allow 'millisecond' pulses of dc voltage (each pulse, say +12 volts dc, but only for a predetermined amount of time). Relatively speaking, the pulses can be wide or narrow (with respect to each other). Energy is only delivered 'during the pulse'; there is 'dead time' between pulses. The wider the pulse, there is less dead time and more voltage gets to the fan - fan goes faster. The narrower the pulse, there is more dead time and less voltage gets to the fan - fan goes slower.

Either method (AC or DC voltage), by nature of the components and the circuit design is inherently precise, and good voltage control would occur to run fans.

These same type circuits can be used by the mobo manufacturer in conjunction with an additional circuit that 'reads' the inside case temperature and runs the fans faster or slower to cause the case temp or the cpu temp to be in a certain range. Most anything circuit wise can be done, but manufacturing cost does become a factor 'based on features offered'.

[chazman113]

You are correct in that the resistor would heat up. Your control 'knob' will 'smoothly' lower the voltage on the fan. The max voltage would probably be +12 volts DC. The voltage could also be an AC voltage equal to that required to run a fan with AC volts; however, probably the ATX standard might require (?) that any ATX form factor computer would have fans run on DC volts only.

If AC voltage, a circuit of 'back to back' SCRs (silicon controlled rectifiers) can 'selectively' reduce the amount of ac voltage to the fan and, therefore, reduce the speed. The AC voltage has the shape of a continuous 'sine wave' provided the SCRs are turned 'fully on' 100% of the time. However, each SCR must 'fire' each 1/2 cycle; an SCR can be triggered so that 'less than a half cycle' is supplied thereby, limiting the 'energy' to the fan and slowing it down. If the SCRs operate 'wide open', the fan gets full voltage and runs at max speed.

If DC voltage, the control knob can allow 'millisecond' pulses of dc voltage (each pulse, say +12 volts dc, but only for a predetermined amount of time). Relatively speaking, the pulses can be wide or narrow (with respect to each other). Energy is only delivered 'during the pulse'; there is 'dead time' between pulses. The wider the pulse, the less dead time and more voltage gets to the fan - fan goes faster. The narrower the pulse, the more dead time and less voltage gets to the fan - fan goes slower.

Either method (AC or DC voltage), by nature of the components and the circuit design is inherently precise, and good voltage control would occur to run fans.

These same type circuits can be used by the mobo manufacturer in conjunction with an additional circuit that 'reads' the inside case temperature and runs the fans faster or slower to cause the case temp or the cpu temp to be in a certain range. Most anything circuit wise can be done, but manufacturing cost does become a factor 'based on features offered'.

Neato. Works for me. Thanks