Hallo everyone,
I had a 12V PSU got broken after unplug violently from wall socket.
Odd....
The schematic resemble this one, exept for resistors values and different MOSFET and DOUBLE DIODE semicons. I've highlighted some paths.
Red: Vcc
Blue: GND
Orange: B(1)+ // B(2)+
Purple and pink: trigger and sense of MOSFET
The issue is:
The putput voltage is floating around 12v when no load is connected (only DMM).
To ensure that is not output caps I added a 1000u. Nothing changed.
Now, I connected a load (12V 0.3A led lamp) and the voltage now ramp up to 10V, sufficient to trigger LEDS and then drop to 8. It rises again to 9.5-10 and drop. The frequency is 200 flash per minute or 3.3 flash per second.
I'm a newbie in SMPS field. The pcb doesnt show any failure component.
Can you help me indetify the culprit?
I would exclude 12V part of circuit. I think the stage before is causing this issue, that then pass trough TRAFO.
Or the octocoupler part.
When Ive removed brutally the PSU from 230V what could be happened?
Thanks in advance
Gabriel
I had a 12V PSU got broken after unplug violently from wall socket.
Odd....
The schematic resemble this one, exept for resistors values and different MOSFET and DOUBLE DIODE semicons. I've highlighted some paths.
Red: Vcc
Blue: GND
Orange: B(1)+ // B(2)+
Purple and pink: trigger and sense of MOSFET
The issue is:
The putput voltage is floating around 12v when no load is connected (only DMM).
To ensure that is not output caps I added a 1000u. Nothing changed.
Now, I connected a load (12V 0.3A led lamp) and the voltage now ramp up to 10V, sufficient to trigger LEDS and then drop to 8. It rises again to 9.5-10 and drop. The frequency is 200 flash per minute or 3.3 flash per second.
I'm a newbie in SMPS field. The pcb doesnt show any failure component.
Can you help me indetify the culprit?
I would exclude 12V part of circuit. I think the stage before is causing this issue, that then pass trough TRAFO.
Or the octocoupler part.
When Ive removed brutally the PSU from 230V what could be happened?
Thanks in advance
Gabriel
Last edited:
Only if the switching time is fast, right?
Since is 3.3Hz and give 10.5V@0.3A (now, hiccup) I expect a still lower freq when correctly working ie. They used 1n4007.
If you say that slow diode could damage the circuit I will swap it. If now, why bother?
Can you explain what aux winding does?
Thanks for help
The aux winding supplies the primary controller IC - after a succesful start.
It is fed by the switiching transformer, thus the diode is exposed to the switching frequency of several 10 kHz.
IN4007 is a no go here - ditch that crap!
It is fed by the switiching transformer, thus the diode is exposed to the switching frequency of several 10 kHz.
IN4007 is a no go here - ditch that crap!
It MUST when you are powering COBs.
Typically no current regulated means lower power output ie. Not pushed to limits.
OT: led philips bulbs in DUBAI have double the filaments than US EU bulb, so they are undervoltage to achieve a super efficiency.
In this case these LEDs are intended to work plung n play. But to confirm, when I supply the TV sat decoder it power on and off like 1 time per second. Pop of audio, tak tak tak tak...
Ok I trust you and I will put a simple UF4004. Is it right? I've also MUR160 but I would save them for other proj.
Ok. If I undestand right the controller IC is feeded initially by the 1Mohm path (3x 360K in series) but then aux path is the lower resitance path and gives current after the first cylce of MOSFET ON/OFF.
Im new to SMPS. By one side, these circuits are smart but I see a lot of fragility in these. If someone has not money/space/weight/consumption issue, then a classic trafo + full bridge rec. + filter caps would be always the best way to go. Am I wrong?
Fine. 2.7ohm.
I traced the PCB. The scheme posted in #1post is not the indentical same of my board.
IMHO the mosfet is fried. I suppose that since you explained the aux circuit.
So the main power start the IC that cant trigger the MOSFET so the inrush voltage and current @300ish V induce the voltage/current in the secondary (lets say 10.5V fews mA). One pulse. Then no more pulses until the caps discharge and the whole circuit thinks to be just powered ON.
Seems logic and coherent with 3.3Hz.
Btw I can hear a tik every time the led got powered. The tik come from MOSFET. Usually the tiks are so fast but also silent so you cant hear a constant freq whine, at least at 50cm distance.
Here I tracked the circuit in the first part of the PCB. LIGHT BLUE is the aux path.
Fine. 2.7ohm.
I traced the PCB. The scheme posted in #1post is not the indentical same of my board.
IMHO the mosfet is fried. I suppose that since you explained the aux circuit.
So the main power start the IC that cant trigger the MOSFET so the inrush voltage and current @300ish V induce the voltage/current in the secondary (lets say 10.5V fews mA). One pulse. Then no more pulses until the caps discharge and the whole circuit thinks to be just powered ON.
Seems logic and coherent with 3.3Hz.
Btw I can hear a tik every time the led got powered. The tik come from MOSFET.
In a normal operating circuit, the tiks are so fast (10-500kHz) but also silent so you cant hear a constant freq whine, at least at 50cm distance. Instead, now, I hear a 3.3Hz tick simultaneously with led blinks.
Here I tracked the circuit in the first part of the PCB. LIGHT BLUE is the aux path.
View attachment 1377506
Rectangles delimit the Qx and ICx.
From top to bottom and from left to right.
Dual diode
431
Opto isolator
6pin PWM controller
MOSFET
GREEN LINE like the scheme, trafo galvanic isolator
I do not believe the MOSFET is fried, because
a fried PowerMOSFET does one final pulse - and then the primary circuit finally disonnects - no pulses following at all
a fried PowerMOSFET does one final pulse - and then the primary circuit finally disonnects - no pulses following at all
