On Sun, 05 Apr 2015 16:21:22 -0400, M Philbrook wrote:

Many switch mode supplies do not like a slow ramp up on the input
voltage. Some cheap designs require it to have voltage sitting there for
proper inrush sequencing to take place for the feed back to get started.

This is a Philips going back a few decades, so I doubt that would apply
in this instance.

Some of these cheap supplies will actually hold the switching mosfet
in a DC state and over head the inductor, power tranny or both.

The chopper in this design is a power TO-3 cased BJT.

Most variacs have only a little more than input side, I don't think
that little over percentage caused the problem.

I'm just thinking maybe if they'd used some caps with a voltage rating of
250V then that would have been briefly exceeded, albeit by a small
amount, by my error. There are some dodgy looking 'tropical fish' caps
that showed clear signs of cracking apart even before this incident!