Switched power supply carnage

[madires]

I have no C/ESR meter, so I changed it with a new one. No success.
But something caught my eye in the Tr. secondary side. If the +5V circuit is dead (D8 or maybe C7/C8) then there is no feedback whatsoever and there is no switching. The Schottky diode looks beefy enough and is acting normal when measured on diodes scale. But those caps... yeah... I definitely have to get a C/ESR meter.

If the 5V rail is dead the TL431 doesn't sink current through the optocoupler's LED. That means, the optocoupler's transistor doesn't sink current. The TNY276's datasheet mentions something about auto-restart in case of an open loop, output overload or output short circuit. If the EN/UV pin isn't pulled down for 64ms, switching will be disabled for 2.5s. Checking the 5V secondary including the TL431 based feedback is a good idea.

[masster]

I give up. I changed the Schottky diode and both capacitors on 5V rail, then the optocoupler. NOTHING. Zero on all outputs.
What have I learned from this? SMPSes are a POS . Ten times more unreliable and fussy than a linear PS. I wasted a lot of time with that little garbage PS. They work like standing in balance on the tip of a needle. I will build myself instead a nice external linear PS, put it in a small plastic box and use a 4 pins connector to feed 5V, 12V, and -22V to DVD. That will give me years of peace.\
Thank you to all contributors.

[nsayer]

Oh, I dunno. If you're going to replace the whole thing, why not design your own switching supply? I'm new to it as well, but from what I see it's not to hard to do it properly.

You might consider going ahead with a nice, beefy +5v supply, and then use switchers to make the other two voltages from that. Do you know what current requirements there are for the device?