I have a small computer SMPS here in parts, it had bad output caps, it's from 2004, possibly never even used. I'm just using it to study. I have made the entire schematic. I put the Vaux/stby part back together, and I can't get that part to start on it's own, not predictably anyways. A picture of just the parts involved is below., the mosfet is inside the TNY267, below too

Applying 5.12V to the real Vstby rail, is exactly what the TL431 needs to see to turn on right and regulate. (LT1009 in the schem, and ZD1 is a 5v0, LTspice is lacking on zeners.)

However, turning on the PSU, the off-line switcher TNY267, only goes into a low power mode. It generates about 4V for Vaux, and barely more than the diode drop on Vstby.
https://pdf1.alldatasheet.com/datasheet-pdf/view/176833/POWERINT/TNY267.html


I have applied 12V/5V/3.3V to the main rails, and they don't have anyway to create more than about 1.2V on the Vstby rail. I don't see anyway the -5V/-12V rails can affect it either, this layout is pretty basic for a SMPS.

But it did start once or twice, without me applying voltage.

I've tried starting it with a 100 and a 22ohm load, that didn't work. That was only on the Vstby, I haven't tried loading the Vaux, the PWM chip is not turned on, I removed a BJT. I better try that too.

As soon as I apply 5.12V on the rail tho, the opto will allow the TNY to go full power, and then it runs just fine, and is not affected by the other rails at all either.

All of the solder has been replaced or retouched on the PCB, again, I took it down to diodes and resistors, and measured the rest with LCR meter.

 

C5 and C6 are 1,000uF, they would have a big in-rush current, IDK if that's vital or not, they both had great ESR

 

Check the values of R% and R6. You will probably need to disconnect one end of one of them to get proper readings.

 

They are both 1.98-1.99Meg, and just carbon 5% type. So I measured the current from them to pin4 Enble/Line-voltage-UV, it's 42uA leaving them, with 2.26V on p4. Disconnecting both resistors, and it starts every time and runs great. There is then 284uA emerging from p4 into the opto. The NPN there stays off at all times unless Vaux goes too high.

The internal regulator is putting out 6.28V, when I 1st had this working it was the proper 5.8V.
In the pdf it says that's normal tho. When the UV resistors are connected and it doesn't start right, it's 4.9V.


I have a model working with the main trans and aux. trans , and I've been using the functions on my scope, and capture events. I got caught up studying it working, and didn't get a good shot of startup events.