I would like to control RGB leds , or other PWN channels , using a micro controlelr, and MOSFets transistors- to push high current/voltage .

I know that MOSFets, are usually not linear-means a vgs above some voltage will open them, with an exponent curve .

My question is , if i set a PWM signal from MCU to a mosfet transistor gate, in order to control its output linearly (rgb leds etc) , does it need to have a certain specs to work, or just setting a PWM signal to any MOSFet gate will do it ?

For example we would like to use this one : http://www.digikey.com/product-detail/en/FDS8984/FDS8984TR-ND/1245054

Does it has the right spec to do it ?

  • \$\begingroup\$ Did you even try to proofread this? \$\endgroup\$
    – Phil Frost
    Nov 30 '14 at 13:06
  • \$\begingroup\$ what do you want? go get a life. I am investing my time in creating new things, not editing thinking all day about this website . \$\endgroup\$
    – Curnelious
    Nov 30 '14 at 16:31
  • \$\begingroup\$ @Curnelious This kind of attitude is not welcome on EE.SE . \$\endgroup\$ Nov 30 '14 at 17:23
  • 1
    \$\begingroup\$ It shouldn't take you a whole day to reread your question, fixing all of the typos, capitilization, punctuation and spelling errors. Remember, you're asking the people here for the favor of free advice, and you want to make it easy for them to help you. If you come here with all kinds of attitude and try to antagonize people, nobody wins. And no, you're not creating anything new; you're doing something that has already been done millions of times over. \$\endgroup\$
    – Dave Tweed
    Nov 30 '14 at 17:31
  • \$\begingroup\$ possible duplicate of Controlling MOSFET and LED from same PWM output \$\endgroup\$ Nov 30 '14 at 19:03

When using a MOSFTE "in saturation" it acts just like an on-off switch.

The higher you go above the gate threshold voltage the lower the resistance of the switch. With a low voltage (0V) applied to the gate, the resistance is massive. Putting a high voltage (say 5V) on the gate the resistance is minuscule.

In fact, from the datasheet for that specific one, at 4.5V the resistance is a mere 30mΩ. The threshold voltage is (around) 1.7V, so there is a region around that area where the resistance changes between (near) infinite and a few tens of mΩ. It is that region that is used when acting as an amplifier - vary the resistance according to tiny changes in the gate voltage. You're going way way outside that area though, so you can completely ignore it.

As you can see from this graph (taken from the datasheet) the current that can flow through the drain rises sharply at the threshold voltage:

enter image description here

Once you pass that threshold voltage you have already lost most of the resistance, the majority being around the 2V area.

All you're really interested in is:

  • Will it be a low resistance at my input voltage?

There are other parameters that are of interest to a more advanced user, but for simple low-frequency switching that's all you really care about.

  • \$\begingroup\$ wow this is a great answer, thank you very much it helped me a lot . But as i said, if you look at 2.2-3v you get a non linear curve, the question is, if this is enough(seems linear to my eye) as an RGB controller, with a 12V DC , that you wants to control from the MCU and cover all colours ? \$\endgroup\$
    – Curnelious
    Nov 30 '14 at 16:36
  • \$\begingroup\$ If i think about it, its not good for most of the MCU's because they drive 0-5V , so ill have to change it, and hurt the resolution of the whole thing . So, is there a good choice to control from MCU with PWM, and high current(3A) ,lets say a led strip ? \$\endgroup\$
    – Curnelious
    Nov 30 '14 at 16:47

I wouldn't drive a MOSFET directly from a microcontroller, there are logic level MOSFETS which work as the name suggest with voltages of around 5V, but generally MOSFETS require 10 or 12V on the gate.

Additionally to switch MOSFETS efficiently you want to be able to supply them with a high initial current to charge the gate capacitance, switching the MOSFET on quickly, you also want to discharge the gate quickly so you need to sink that charge. This is a simplistic way of looking at it, and there are plenty of in-depth articles online.

The datasheet for your device shows a graph with the Drain to Source resistance, against Gate to Source Voltage. At 5V Gate Source Voltage (at 25C) the Drain to Source Resistance is 22.5mOhm and at 10V it's 19mOhm, not much but it obviously has implications on the efficiency, as well as how hot the MOSFET will become pending on how much current you are switching. If your drive voltage is below 5V then this relationship becomes much more critical, with the FDS

I mainly use MSP430G microcontrollers, these use 3V and have an approximate maximum current per GPIO of ~6mA. A circuit I tend to use is a totem-pole pair of transistors, usually a BC337 and BC327, the circuit can be easily found on the internet. To drive the totem-pole pair I use a BC550, this does invert your PWM signal so be aware of that. Alternatively I have also used the TC1413N, high speed MOSFET driver this works well.

To provide extra current to the FDS8984, you could simply add the totem pole pair between your MCU and the FDS8984, this should help improve the efficiency of the circuit as opposed to just using the MCU as a driver, as well as offering over current protection to the MCU.

  • \$\begingroup\$ Thanks a lot for your help! well, i am little bit confused, i would like to be able to control a led strip , so first it has to drive a lot of current, second, it needs to be able to change from 0 to 5/12 v , to cover all colours of the leds. So ,this is why i chose this one to be driven by PWM, and into a led strip. did i chose wrong ? \$\endgroup\$
    – Curnelious
    Nov 30 '14 at 16:42

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.