LED Blinking EMI

Thread Starter

EE_1

Joined Sep 6, 2013
2
Hey guys,

I'm wondering if its okay to route simple led traces over a plane split in a PCB design. We are talking about CPLD LVCMOS edge rates. Something as simple as a blinking LED indicator strikes me as a non-critical, route anywhere type of thing. However, if one was to blink such a light, would significant EMI be produced given the edge rate?

I really would like to avoid such considerations for led signals because trying to avoid plane splits for everything drives up the number of PCB layers needed.

I appreciate your feedback.

Thanks,
EE_1
 

MrChips

Joined Oct 2, 2009
30,720
It depends with what it is likely to interfere.

Anytime a digital signal switches it will generate some noise in adjacent circuits and circuits connected to the same power supply. You can minimize the noise on the power supply by paying attention to trace routing, width, supply and ground loops, ground planes and decoupling capacitors.

It would also help to reduce the current to the LED. If it is a critical situation you might attempt to slow down the rise and fall times of the signal to the LED.
 

Thread Starter

EE_1

Joined Sep 6, 2013
2
Hi Mr. Chips,

Thanks for the quick reply. I agree, proper IC decoupling and grounding is a must.

I'm just wondering how bad the EMI would be from the current loop that would arise from such a situation. The current has to find a way across the split, which would be the closest decoupling cap. Large current loops create EMI.

But it is just an LED signal in question. We are talking about frequencies well under 60Hz, because beyond that the human eye cannot see the blinking, well actually, even at 60Hz you can't see the blinking:)

But the LVCMOS edge rates make me worried. But since I'm not switching at 2.5GHz, do any of you really think such EMI would be a problem?


Thanks,
EE_1
 

MrChips

Joined Oct 2, 2009
30,720
You have to tell us more about what the rest of the circuit is doing.

If everything is digital you should not have a problem. The problem arises when you have a mixed signal application, both digital and analog in the same system.

I am currently working on such a system digitizing a low level analog signal while simultaneously generating VGA horizontal and vertical sync signals. If I am not careful I would be able to see the sync signals in the digitized waveform.

It is not just about the frequency, whether 60Hz or 2.5GHz. It is the edge of the signal that causes the problem. Every transition from low-to-high and high-to-low has the potential to generate frequencies all the way to infinity. One solution is to slow down the rise and fall times of the transitions.
 

WBahn

Joined Mar 31, 2012
29,979
If you have a current limiting resistor in series with the LED, then that is going to slow your edges. You can always put a small capacitor or T-filter to slow the edges more, though this is probably overfill.

Also, you can use stitching capacitors to AC couple the various planes so that image currents can stay close to the signal currents thus reducing the effective size of the EMI radiating loop.
 
Top