The Original power supply blew taking out both High Power 24N60CFD Mosfet.
Which in turn took out the 3watt bias resistors and a 1/4 watt resistor which I have no idea what value it was.
You can see in the photo that the left Mosfet left most leg has blown, and various resistors on the Photo.
Because of your are interest, opened the S-500-24 shown below.
Nothing is visually burnt measured the Fuse and Thermal fuse both in good shape.
Connected power and measured power threw the fuses to the pi filter, AC power arriving and measuring on the full wave rectifier B02.
Changed over to DC voltage Voltage on B02 rectifier measures 330VDC All high voltage caps are good still no output voltage.
Measure random spot to see where the voltage starts and stops. Opps! Whoa!!!! what happened power supply came to life.
Well that exciting and was very short lived, first it jumped into a powered on state now it is back to not working again.
Well the diagnostics are completed and verdict has been made the supply has a cold solder connection.
With the this new known information, completely dissembled the supply and removed the printed circuit board from case.
Then completely examined the board with no visual detection of where the cold solder joint is, all connections look good.
OK. Well at this point in time nothing to loose, only left with one choice, reverted to the shotgun method solder everything.
Connected the Weller PU120 soldering station, dragged out my .022in QQ-S-571 solder and began solder every connection.
Reassembling the Supply making sure that all insulation material is between the heat sinks and TO-3 packages.
Connected the power this is where I wish I had a variable AC supply then I could adjust the power up slowly.
I feel good this is going to work, Time to see if my labors worked or will there be a loud bang and a huge puff of smoke.
And the verdict is in the supply is running