[Info-vax] ChatGPT solved it for me, again...

[gah4]

On Wednesday, February 15, 2023 at 5:56:21 AM UTC-8, Steven Schweda wrote:

(snip on "too much power")

> Not really. Some (mostly older) voltage regulators had a minimum 
> load requirement. Connect them to a smaller load, and the output 
> voltage could exceed the rated value, with potential damage as the 
> result.

When the IBM PC/AT came out, the hard disk was optional.
If you didn't order it, you got a 15W resistor where it would go.

Sometime later, I had such a resistor that I would use when 
testing PC power supplies.

And I did once have a non-IBM power supply, as they say:
"let the magic smoke out".

As far as I know, the voltage didn't exceed the rating,
but after the smoke came out, I didn't try again.

ATX power supplies are required to work under 0 load,
They might not be required to work well, though.

I have an ATX power supply tester that tests with
zero load.  Of course that doesn't really test it!

In any case, yes, using too big a power supply
isn't a good idea.  And even when it works, they
are less efficient. (Or, at least, the efficiency
requirements are lower.)