Power suplly resets itself in nominal input voltage-URGENT

Is the power supply shutting down and staying off?

If so, I would look at your line sensing OV/UV circuitry as a possible culprit.  You could disable this circuitry and see if the problem goes away or continues.

If, however, the power supply is going into auto-restart (the power supply will switch for a short while then shutdown, switch for a short while, then shutdown, repeatedly) then this indicates that the controller is not receiving a proper feedback signal on the control pin.  You could try adjusting the value of your soft finish circuit or eliminating.  It probably wouldn't be a back idea to verify good choices of your feedback circuitry components as well.  Does it do this under a particular load condition?  Maybe that can point us in the right direction.

-The Traveler

Dear Traveler,

In 220VAC input The power supply is going into auto restart mode in no load case. In fact if I increase the load, the problem still exists.

But if I increase input voltage slowly from 0 to 220 V in this case it studies well..

Fınally for auto-restart condition.. If I unplug the Power supply from main electric grid it starts to work for a few seconds(in no load case so the capacitors does not discharged) ..

The strange problem is we produced and sold the same Power supply, about 5000 pieces,  with the same feedback and soft start circuit, and it was the first time I see such problem... 

I will try to change soft start circuit..

Best regards, 

Kemred -

Do you have a circuit schematic or Pi Expert file I can take a look at?

If this were my design, the first things I would do are:

-Eliminate any of the "extra" features (UV/OV, OVP, Soft finish, etc) and try to get the device as close to 3-terminal operation mode as possible.  Add the features back one by one and see if you can track it down to a particular section of the supply.

-With the power supply going into auto-restart, this indicates that a proper signal is not getting back to the control pin of the IC by the end of the IC's softstart time period (17ms).  I would check output voltage waveforms during startup to see how the PSU is behaving during startup with different line voltage inputs.

-If this power supply is already in production and the design hasn't been modified, then *something* has to have changed.  A part substitution, a new batch of transformers, a different lot# of capacitors, etc.  If you're working with a CM, they should be able to help provide this information. 

-The Traveler

Dear Traveler,

I find the problem, in production the snubber TVS and snubber ultrafast diodes for the primary are connected, reversely. so The circuit behaves like TVS paralel to transformer...

I thing with this setup whan I applied 220V to the input the circuit behaves like OVP... or something similar to it.. when we correct the directions of these TVS and UF diode.. the circuit behaves normally as accepted..

But I don't understand how the supply works when I increase the input voltage slowly and don't enter OVP even high voltage..  

My best guess is that, at low line, the primary could still process enough power (current is bypassing the primary through the 220k resisters) to have the output reach regulation.  At high line, the current through the 220k resisters is probably enough to make the current sense circuitry hit Ilim very fast, not allowing the output to reach regulation.  I'm honestly suprised that it even worked at low line. 

-The Traveler