LED Strips PWM control

Hi all.

I'm currently trying to construct a board, where I can wirelessly control an LED Strip using my phone's Bluetooth.

I'm using an atmel 328PU microcontroller for the main control of the system.

What I'm trying to achieve, is to have up to 5 strips attached to the board. Each has to be able to be turned on/off individually. The PWM control of each strip is controlled together, mainly because of the lack of PWM output pins.

So far I've tried to design the on/off mechanism using MOSFET transistors. My main problem here, is that the MCU only provides 5V output, which is rarely enough to fully turn on most of the MOSFET's. This cause a huge heat dissipation, which is not the intention.

I have a decoder from china, where you can control the led strips using an IR remote. From what I can see, they use Atmel 125 24C02 for controlling and VS30 diode rectifiers for the pwm control. This seems to be working fine, without any uncontrolled heat dissipation. However, I'm not quite sure this will work in my application?

My question is, is there any easier way of controlling the on/off and pwm mechanism than using MOSFETS, or a transistor as a switch in general?

Each STRIP use a maximum of 14 W and 1A. I have 12V and 5A Power supply.

You should use "logic level" Mosfets when the gate gets only 5V. There are hundreds of Logic Level N-channel Mosfets available.
For example an IRF540 is an "ordinary" old N-channel Mosfet that needs a gate-source voltage of 10V to fully turn on.
The newer IRL540 (see the "L"?) is Logic Level.

Audioguru said:

You should use "logic level" Mosfets when the gate gets only 5V. There are hundreds of Logic Level N-channel Mosfets available.
For example an IRF540 is an "ordinary" old N-channel Mosfet that needs a gate-source voltage of 10V to fully turn on.
The newer IRL540 (see the "L"?) is Logic Level.

Click to expand...

THanks for the answer.

I happened to come across a logic level MOSFET K4033, which has VGC of only 4V. However, I'm still struggling getting enough through the transistor to drive the leds proberbly.

At the moment I'm having a 10k Resistor from the power source to the Collector. I'm aware it limits the current a lot, which is also the reason for the loew current flowing through the transistor. But if I use a lower resistor, like 1k, it simply gets fried instantly.

I'm taking the output (vout) right after the resister at the collector pin.

But if I use a lower resistor, like 1k, it simply gets fried instantly.

I'm taking the output (vout) right after the resister at the collector pin.

Click to expand...

It's hard to believe that 1k would allow so much current that something gets fried.

Are you using the typical way to power a load with an N-mosfet (or NPN transistor)?
Like this:



The resistor can be reduced to 12 ohms. Then your led strip will get 1A maximum.

LED Strips are designed for car voltages 11.5 to 14.5V with current limiting resistors

THus at 14V they take 14W with 1A.


You could drop up to 250mV with an RdsOn of 250mOhm or 1/4W @1A and the brightness difference would not be noticeable, but lower is better.

e.g. www.Digikey.com search MOSFET or
> Product Index > Discrete Semiconductor Products > FETs - Single
Results matching criteria: 35,009 on Digikey

>locate FET Feature> choose Logic drive 4.5V then> Mounting type (Through Hole?) ... SMD are better> then [x] in stock then [ search] then click by price descending (^) on results.

For under $1 @1pc https://www.digikey.com/product-detail/en/FKI10531/FKI10531-ND/4515097
ID = 11.9 A, VGS = 10 V 34.7typ 54.5 mΩ max
ID = 6.0 A, VGS = 4.5 V 36.5typ 56.5 mΩ max


You can save a few pennies but you wont need a heatsink with this choice. I prefer SMT FETs if you are laying out a board.

So I've tried to use a similar transistor as the one you replied with. I found out that my load resistor was causing the heat problem. My attempt to a solution is:



So I have 2 RBG LED's, where the first MOSFET controls the ON/OFF state and the second three RBG control each of the R, B and G GND. The second MOSFETS are controlled using PWM.

The problem now is that the first MOSFET is always open. I am aware that this is because of an n-channel with a load at its low state. Using a P-channel MOSFET should solve this problem. However, I only have N-channel MOSFETS available, and was interessted to know, if there is a safe and possible workaround? It would take a month to order some P-channel MOSFETS, so a temporary solution would be the most prefered choice

santais said:

I happened to come across a logic level MOSFET K4033, which has VGC of only 4V. However, I'm still struggling getting enough through the transistor to drive the leds proberbly.

Click to expand...

Maybe you are takling about a Japanese 2SK4033 Mosfet that is not recommended for new designs.

At the moment I'm having a 10k Resistor from the power source to the Collector. I'm aware it limits the current a lot, which is also the reason for the loew current flowing through the transistor. But if I use a lower resistor, like 1k, it simply gets fried instantly.
I'm taking the output (vout) right after the resister at the collector pin.

Click to expand...

A Mosfet does not have a collector.
We have no idea how you connected it so please post your schematic.

4V threshold MOSFETs migrated to 2.5V threshold with some now at 1V and going lower.

If the MOSFET is hot, get one with a lower RdsOn. If the R is too hot, use Ohms Law and choose a better Source voltage, string size, voltage and R drop

Audioguru said:

Maybe you are takling about a Japanese 2SK4033 Mosfet that is not recommended for new designs.


A Mosfet does not have a collector.
We have no idea how you connected it so please post your schematic.

Click to expand...

Yeah sorry for the mistype. Meant Drain and not collector. You can find a schematic below. The schematic is incomplete and only meant as a sketch of the overall schematic. If it's still too unclear, let me know and I try to redraw it



So each pair of RBG LED is connected to the same MOSFET PWM Drivers. Only the ON/OFF state of each RBG LED is the one I wish to control independently of the ON/OFF state of the other RBG LEDs.

- - - Updated - - -

SunnySkyguy said:

4V threshold MOSFETs migrated to 2.5V threshold with some now at 1V and going lower.

If the MOSFET is hot, get one with a lower RdsOn. If the R is too hot, use Ohms Law and choose a better Source voltage, string size, voltage and R drop

Click to expand...

Got my hands on a MOSFET with Rds on 0.07 so my heating dissipation problem has now been solved. Thanks!