Man-wai Chang ToDie (+MS=V32B) said:
I meant using it as the power switch for a typical PC (ATX) ....
This one:
http://www.allproducts.com/manufacture9/rongfeng/rf112b.html
The S732 here will do the job.
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=360-2095-ND
http://www.nkkswitches.com/pdf/stoggleshighcap.pdf
It is a real switch, and can handle the surge rating
of the ATX supply. The switch costs $50, and for
that price, you can buy an ATX supply with its
own switch included
If you're going to use a circuit breaker, it will
have a lower limit on the number of ON-OFF cycles.
The S732 is rated for 25,000 operations. You can
switch the computer 10 times a day, for the next
6.85 years, and the switch contact will last that
long. The circuit breaker will have a much lower
cycle rating (get the datasheet from rongfeng and
see for yourself - I don't see any data on their
website).
Now, some in the audience will think that my choice
is overkill. Maybe. But we had an embarrassing
incident at work, involving this very subject. A
computer we built, used a rocker switch on the front
panel, to provide control of an open frame power
supply inside. The switches began to disintegrate
(physically fell apart) after a year of usage. I had
one fall apart in my hands one morning (because we
used the computers we built, for our own desktops).
While the switch had a continuous rating that seemed
appropriate for the 250W supply, it was not appropriate
for the surge rating of the supply. The switches
were destroyed by the surge each day the switch was
used. A later redesign involved the use of a 30A
relay, which permanently solved the problem (with
a greater expense).
This is the surge rating of a certain Enermax power supply.
"Inrush Current Limiting
80A/230V and 40A/115V max. during cold start"
That current only flows for a couple cycles of the AC power,
until the main capacitor is charged. In the supply, a
thermistor is used, which present a high initial resistance,
and when it heats up, the resistance drops, so that the
efficiency of the supply is minimally affected. In the
circuit diagram here, NTCR1 in the upper left hand corner,
limits the inrush current. But the value of the inrush current
is still high using that method, and for an inferior quality
of switch, the switch can be destroyed, as we discovered
for our product.
http://www.pavouk.org/hw/en_atxps.html
It would be really nice, if all the switches had a surge
rating, because then you could tell exactly which switch
was good for the job. There is no surge rating in the above
datasheet, and many switches won't give an indication of
any surge capability. If the continuous rating is high
enough, then you wouldn't have to worry. My guess would be,
a quality 30A switch is probably good enough.
Some ATX power supplies may have a lower inrush current
value, because some of them (perhaps the ones with active
PFC), may implement inrush limiting with the active circuit
instead. The ones using NTCR1, are more brutal, and
can do more damage to a mechanical switch. But the NTCR1
method is cheap, which is why it is used in ATX power
supplies. And it is difficult just looking at a switch,
to guess how well it will handle the initial current
surge.
Have fun,
Paul
