6
\$\begingroup\$

I'm currently planning a project on a low-cost WiFi-enabled RGB LED controller and could use some help in looking over my parts list and general concept.

What I'm wanting to do is a build a relatively low-cost, WiFi-enabled RGB LED (strip) controller. In the end it should be controlled via an Android App which is already being developed in my free time. For the WiFi module I want to use the infamous ESP8266 board, namely an ESP-07.

We are talking about normal, RGB, LED Strips (5050 SMD in my case) - nothing fancy.

The plan is to flash it with nodemcu firmware (can't provide a link, but googling "nodemcu" will show what it is about, but it doesn't really matter for the sake of the question), write a simple TCP server implementation which allows to manipulate 3 GPIO pins' duty cycle over the network. I have no problems related to the software or the programming of either the MCU, the App or anything else. What I'm unsure about are the electronics involved here. I've come to understand that I want to use transistors or MOSFETs to drive the LED Strips 12v using the 3.3v logic of the ESP8266. From what I know, I should choose a FET that can be switched fast (requirement for PWM) and has a maximum gate threshold voltage below 3.3 Volts. Based on these requirements and the availability at my supplier I've chosen the IRFZ44N and I want to wire it directly to the three GPIO outputs of the Microcontroller.

Is there any reason for this being a bad idea? Should I maybe add a diode? I don't think so since LEDs are resistive load and not inductive (so there shouldn't be voltage spikes or any of the likes, but I'm a layman when it comes to electronics, far more involved in programming..) Please tell me if there are any pitfalls to this plan and thank you for taking the time to read through my plan.

\$\endgroup\$
  • \$\begingroup\$ Are we talking individually addressible LEDs, or the ones where you can only control the entire strip as a single entity? What's the maximum current draw of one channel of the strip? \$\endgroup\$ – Nick Johnson Jun 28 '15 at 10:48
  • \$\begingroup\$ I don't have the datasheet for the strip at hand but it's a 5050 SMD strip and it runs off of a 5A power supply for about 5 meters of the strip comfortably. Could probably add some more, so the maximum current per channel would be something along the lines of 5A divided by three, for which the MOSFET seems plenty. Edited the original post, too. \$\endgroup\$ – Robin Heller Jun 28 '15 at 10:55
  • \$\begingroup\$ In case someone pass here like I did searching a mosfet to control strip from the esp8266 you can use the IRLZ44N. Source: www.esp8266color.com \$\endgroup\$ – RiRomain Jun 9 '16 at 20:31
  • \$\begingroup\$ @NickJohnson, the 12V LEDs are typically not individually addressable. These are the ones with 4 control lines, Red, Green, Blue and Ground. The individually addressable LEDs typically run on 5V and have 3 control lines, +5V, Ground and Data In. \$\endgroup\$ – linhartr22 Sep 6 '17 at 15:43
8
\$\begingroup\$

The FET you link to is not suitable for your application. Taking a look at the "typical characteristics" table, we see this:

enter image description here

Vgs(th) is listed as between 2 and 4 volts - already it could be higher than our supply voltage, but not especially the "conditions" column on the right: the specification is given with only 250 microamps of current. Your LEDs, presumably, will draw a lot more than that.

To figure out if the FET is suitable, you need to know how much current you plan to draw, then look for the graph plotting "typical output characteristics":

enter image description here

Look for a line corresponding to the gate voltage you plan to use - in this case, there isn't one low enough - and look for where the curve flattens out. If that's at a current level greater than the amount of current you need to supply, you're good. Alternately, look at the "typical transfer characteristics":

enter image description here

This plots maximum current as a function of gate voltage, so you can trace a line from your MCU's gate voltage - 3.3 volts - to determine what the maximum current level at that gate voltage will be. Again, in this case, the plot doesn't actually go low enough to evaluate that on this particular FET.

Keep looking - there are FETs that meet your requirements - and remember that everything depends on how much current you plan to conduct.

\$\endgroup\$
  • \$\begingroup\$ What about using a second, small transistor to switch the Mosfet? 3.3v Logic -> small Transistors Gate (12V on Collector), small Transistors Emitter -> MOSFET Gate (12V again on Collector) - does this make any sense? \$\endgroup\$ – Robin Heller Jun 28 '15 at 11:14
  • \$\begingroup\$ @RobinHeller Yes, that's possible - but selecting a power MOSFET with a lower Vgs(th) would be much more straightforward. \$\endgroup\$ – Nick Johnson Jun 28 '15 at 11:15
  • \$\begingroup\$ What characteristics would I need to look into to find a regular transistor compatible for switching the FET? I can get the IRFZ44N quite cheaply from China so it is more of an economics decision. \$\endgroup\$ – Robin Heller Jun 28 '15 at 11:19
  • \$\begingroup\$ Same as the above, only the current you'll need to conduct is much lower. Or, use a low power BJT. I find it difficult to believe that the IRFZ44N is the only FET you can get cheaply from China! \$\endgroup\$ – Nick Johnson Jun 28 '15 at 16:21

protected by Dave Tweed Nov 28 '16 at 14:35

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

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