0
\$\begingroup\$

I have a 5VDC fan (AUB0505HD), and my ultimate goal is to drive this fan using a 3.3V PWM signal from a Raspberry Pi Zero.

I haven't worked with transistors much outside of EE school, so I don't fully understand how to use them practically. I did some reading, and I came to believe that a MOSFET should help me accomplish this. I went to Digikey and selected a MOSFET that I thought would do what I wanted: ZVN4206AV. I selected this MOSFET because the fan datasheet states that it will need approximately 330 mA at 5V. The threshold voltage on the MOSFET is 3V max, which I thought meant it would fully conduct at 3V, which is perfect because I'll have a gate voltage of 3.3V. As it turns out, after testing, I learned that this is not the case -- the 3.3V gate voltage is only enough to get maybe 1/3 power out of the fan.

Clearly I misunderstood some of the MOSFET specifications. Now, I'm at a loss regarding how I can get the maximum power out of the fan using my setup. Here's my example circuit:

circuit

I need the 5.5V regulated power for the Raspberry Pi Zero to operate, so I was hoping to piggy-back off of the regulator on the fan as well. Is this possible, or do I need to step up to a higher voltage for the fan since I'll have voltage drop on my switching components?

I looked at BJTs, and it looks like they will give me a voltage drop of maybe 1V across just the BJT, which is undesirable... Maybe I don't understand something with those as well.

Any help is appreciated.

\$\endgroup\$

3 Answers 3

Reset to default
0
\$\begingroup\$

Your approach is fine. Here are a few pointers.

  1. The chosen MOSFET has higher resistance at the voltage you are driving. The datasheet is your friend. The device just starts conducting at that range. enter image description here

  2. There are no shortage of MOSFETs which had lower \$V_{GS}\$
    enter image description here

TSM 2314 for example.
3. Place a diode across the fan connection for fly wheeling the reverse voltage which will be generated when you cut the power to the fan suddenly. The fan might have already one inside it, unless sure you can place one outside too. Look for free wheeling diode.

\$\endgroup\$
7
  • \$\begingroup\$ Thanks! I just went back on Digikey to find transistors with better characteristics with a low Vgs. I appreciate your insight! I'll look into diodes for the fan... any particular characteristic to look for on those? \$\endgroup\$
    – Joshua Granger
    Feb 22, 2020 at 10:26
  • 1
    \$\begingroup\$ @JoshuaGranger the current rating of the diode should be atleast the actual load current. The voltage rating has to be say twice the supply or even a factor times more. 1N4001or the anyone in the series are just fine for this application. \$\endgroup\$
    – User323693
    Feb 22, 2020 at 10:32
  • 1
    \$\begingroup\$ How do you know if the fan’s internal cap for the Hall commutation pole switch can handle the ripple current of PWM? Why do you approve/suggest or need PWM? \$\endgroup\$
    – Tony Stewart EE75
    Feb 22, 2020 at 16:17
  • \$\begingroup\$ The fan us an active switched coil with low ESR cap. Why do you presume that needs a Flyback diode? Bad advice. -1 \$\endgroup\$
    – Tony Stewart EE75
    Feb 22, 2020 at 16:19
  • 1
    \$\begingroup\$ \$V_{DS}\$ need not be large enough. Once the gate source voltage is high enough MOSFET will be like a small resistor with a small resistance. The voltage will be current times the resistance \$\endgroup\$
    – User323693
    Feb 23, 2020 at 17:57
1
\$\begingroup\$

In order to reduce , temp rise , power loss , voltage loss and resistance of the switch, it’s current rating needs to be at least 5x your load for good performance. It must also be a low logic voltage capable switch such as Vt=1V.

It you chose wisely , you should not need a heatsink, but may need to consider the far great choices of SMT or get a TO220 size part.

Also since there is NO thermal control stated, you do not need PWM and it will operate near 2/3 rated RPM and current as you using 3.3/5V and fan load currents are fairly linear with voltage.

\$\endgroup\$
5
  • \$\begingroup\$ Thanks for the comment. Are you saying the MOSFET current rating should be at least 1.5A? I didn't realize that. The application of this circuit will be for thermal control on something, yes. That's why I'm trying to get it working with PWM :) \$\endgroup\$
    – Joshua Granger
    Feb 22, 2020 at 10:35
  • 1
    \$\begingroup\$ 1st your load should be closer to 2/3 of 330mA. 2nd it needs a switch rated for 5x that to be able to operate without a heatsink and 3rd that means your voltage drop is acceptable. 4th look for RdsOn << 0.1 Ohm at 3.3V \$\endgroup\$
    – Tony Stewart EE75
    Feb 22, 2020 at 10:41
  • 1
    \$\begingroup\$ I gave you 4 answers not a comment. \$\endgroup\$
    – Tony Stewart EE75
    Feb 22, 2020 at 10:42
  • \$\begingroup\$ I appreciate it! I've upvoted your responses as they were very helpful :) \$\endgroup\$
    – Joshua Granger
    Feb 23, 2020 at 2:16
  • \$\begingroup\$ Although a BLDC motor + fan each have nonlinear current characteristics, the combination equiv.average current is fairly linear like a resistor.with an LC filter in-spite of commutation noise \$\endgroup\$
    – Tony Stewart EE75
    Feb 24, 2020 at 18:45
0
\$\begingroup\$

It's BLDC, it's not controlled this way because it has internally three phases. Usually the fan comes with a smal controller that has pwm input. Your way you should have a very low switching frequency to let it start and stop, this is definitely not a way to control a fan.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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