1
\$\begingroup\$

I have a PC fan controller that is capable of powering 10w per channel. I'd like to put 23w of fans on a single channel, so I am trying to find a way to use the electronics already on the fan controller to activate a MOSFET or other transistor that will be able to handle 23+w.

Basically, the theory I thought would work is to look at the PNP transistor on the fan controller, and take a wire from the gate of that transistor and using it as the gate on the MOSFET to make the MOSFET switch quick enough to step the voltage and vary fan speed.

Unfortunately, doing this did not result in what I wanted to do. The fans spin at a constant rate of approximately 60% top speed. I only have one fan tachometer wire wired to the feedback of the fan controller, so it doesn't get any weird signals from out of phase tach signals.

Is there any good way to accomplish this?

Here is the equipment I am using to test this with.

Fan controller: [BitFenix Recon]

PNP Transistor in fan controller: B772 PNP Medium Power Transistor

MOSFET: ST P14NK50Z N-Channel 500V MOSFET

\$\endgroup\$
  • \$\begingroup\$ Do you have a link to the Fan Controller datasheet, and a schematic of your test circuit? \$\endgroup\$ – helloworld922 Jul 1 '13 at 19:32
1
\$\begingroup\$

I'm taking a guess that the B772 (PNP) collector is connected to the +12V terminal of the fan and that the circuit switches this on and off as a PWM control of speed.

To replace the NPN transistor use a P Channel MOSFET. This needs to be switched ON by grounding the gate so an inverting transistor is needed.

enter image description here

When the B772 is turned ON a small current flows through R1 and R2 and Q1 is turned ON pulling the gate voltage to about 0.1V. This turns the MOSFET ON this supplies current for the fan. There are lots of MOSFETs out there that would be suitable.

Be careful you don't overload the power supply of the controller

\$\endgroup\$
  • \$\begingroup\$ Thanks, I will try these solutions out tomorrow and see what happens. \$\endgroup\$ – Michael Dornisch Jul 2 '13 at 5:27
1
\$\begingroup\$

100k is way too much resistance for reliably driving the gate off for a MOSFET doing PWM. It might work OK with a 1k pull-up (I'm talking about R3.) Better would be to use a dedicated MOSFET gate driver.

However, you could likely replace the entirety of R1/Q1/R3/Q2 with a single MOSFET N-channel low-end switch. Hook the fans between +12V and the drain of the MOSFET. It would be equivalent to your circuit. Again, for best gate drive control, try a smaller resistor, such as 1k, for pull-down.

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
  • \$\begingroup\$ Thanks, I will try these solutions out tomorrow and see what happens. \$\endgroup\$ – Michael Dornisch Jul 2 '13 at 5:28
  • \$\begingroup\$ @MichaelDornisch I didn't write the question I simply proposed my answer to the question asked. Your answer seems to assume that my answer doesn't work. Fans, as far as I know, are usually controlled by connecting them to the 12V line through some form of switch (transistor or MOSFET) with the ground common. The original question solution was to use an N channel MOSFET with the gate connected to the base connection of the B722 PNP transistor which didn't work. Your solution connects to the emitter and should work. Both answers have their merits. \$\endgroup\$ – JIm Dearden Jul 2 '13 at 12:15
  • \$\begingroup\$ @JIm Dearden: Your solution will likely not work when driving the gate using PWM. It will likely keep the MOSFET in the half-on zone, overheating and killing it. The reason for that is the way-huge 100 kOhm pull-up resistor, as I suggested. \$\endgroup\$ – Jon Watte Jul 2 '13 at 18:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.