So, you have this wonderful new design and it almost meets its EMI requirements, but not quite. To your consternation, there is just enough conducted EMI to violate your required limit and that EMI is at least partially coming from a variable brightness LED.

The variable brightness is being controlled by pulse-width modulation (PWM) of the LED current and the resulting pulse current waveform is your culprit. You need the PWM control, but you need to suppress the EMI it's generating.

The following sketch shows a way of doing this.

Figure 1
EMI suppression of PWM controlled LED

The circuit in yellow on the left is self explanatory. As Q1 turns on and off, so does the LED current but the current drawn from the +5V rail is changing up and down too and making EMI.

The circuit in blue on the right adds a constant current source of Q3 and Q4 whose current is steered through the LED and then not steered through the LED. As before, the duty cycle of the pulsed current through the LED controls the LED brightness, but the current drawn from the +5V rail is held virtually constant with very little generation of EMI.

True, there is a phase inversion difference between these two circuits and the driver on the right uses a lot more parts and uses up more power than the driver on the left, but it you can afford those costs, the reduced EMI benefit can be worth it.

John Dunn is an electronics consultant, and a graduate of The Polytechnic Institute of Brooklyn (BSEE) and of New York University (MSEE).

Related articles: