LED Drivers interfere with each other.

[eem2am]

Hello,

We have two 8W , Switch mode LED Drivers on the secondary side of an offline flyback. 9double sided PCB)

The PCB is very heavily populated and we don't have a ground plane as such.
The PCB ground tracks are very thin and sparse.

-though we have implemented pretty much the following grounding scheme............

PCB Grounding scheme:-
https://i48.tinypic.com/nl7bt2.jpg


anyway, we are getting "beating" in the LED current when both LED Drivers operate simulataneously (they are not synchronized but switch at near 200KHz)


Beating in the LED current:-
https://i47.tinypic.com/2wehp8i.jpg

.....beating is the blips every ~1.3ms or so

do you know why we get these beating phenomena?


Incidentally, the voltage of the ground pads of C1 and C2 is varying in relation to each other.....is this OK?

Here is the voltage of C2 ground pad with respect to C1 ground pad.....

Voltage between ground pads of each LED driver's input caps:-
https://i50.tinypic.com/ztggw9.jpg

Is it a problem that the grounds of these two caps (the input caps to the two LED drivers) is not at the same potential, in this way?

 

[keith1200rs]

I am not sure the capacitors across the LEDs are a good idea, although not likely to be the cause of your problem.

Also, your regulator input capacitors are rather small.

However, I suspect the problem is the grounding/tracking given your description. You can probably prove that with some braid or copper tape to beef up the tracking on the prototype. For those sort of currents you really need some very wide copper - look at the applications note for the LM3406 evaluation board (AN-1750) to see what is normally used.

Keith.