LED leakage current via earth

[yassin.kraouch]

Hi,
I have a power source that generates 50V/70hz square wave to drive LEDs( there is full bridge rectifier and LED driver), mounted on an FR4, and this drives a LED PCB which is mounted on aluminium PCB, in the input of the circuit there is 15nF between the input and the metal chassis, when the frequency change to 100hz the microncontroller should switch OFF the LED board, and stay OFF, the problem is when the LEDs are OFF, there is a leakage i think is coming betwee, the line and earth and goes to the LED PCB, any idea how to solve this issue ?

 

[KlausST]

Hi,

for a current to flow you need a closed loop.

I can´t imagine where this can be.

Show your complete circuit including all earth GND connections.

Klaus

 

[yassin.kraouch]

Hi,
I have a power source that generates 50V/70hz square wave to drive LEDs( there is full bridge rectifier and LED driver), mounted on an FR4, and this drives a LED PCB which is mounted on aluminium PCB, in the input of the circuit there is 15nF between the input and the metal chassis, when the frequency change to 100hz the microncontroller should switch OFF the LED board, and stay OFF, the problem is when the LEDs are OFF, there is a leakage i think is coming betwee, the line and earth and goes to the LED PCB, any idea how to solve this issue ?

difficult to show all schematic, as we have two separate boards, one for the driver, the other one is for LED board, i am 100% sure there is a closed loop, between input line and LEDs' ( as LEDs are on aluminium PCB) via the 15nF capacitor that connect the line to the chassis, so my problem is how to solve this issue.

 

[KlausST]

Hi,

difficult to show all schematic

... then it will be difficult to help.

even a hand drawn sketch would do.

Good luck.

 

[yassin.kraouch]

i try to draw something, , is this ok ?

 

[andre_luis]

The above diagram is meaningless in the sense of not informing the current path on the LED.

 

[KlausST]

Hi,

"full bridge and filtering"
* is isolated between input and output?
* output side is somehow (capacitor?) connected to line input or EARTH_GND?

Klaus

 

[yassin.kraouch]

The above diagram is meaningless in the sense of not informing the current path on the LED.

i do not agree with, as you clearly see the current path going from the source to the chassis via the capacitor, and then goes from the chassis to the LED's and from the LED's back to the source.

- - - Updated - - -

Hi,

"full bridge and filtering"
* is isolated between input and output?
* output side is somehow (capacitor?) connected to line input or EARTH_GND?

Klaus

no isolation,
no capacitor in the output side, but i think there is parasitic capacitance between the LED's and the aluminium substrate.

 

[andre_luis]

then goes from the chassis to the LED's

You should not connect the Earth to anything on the circuit.
It is hard to guess what is happening inside the box.