R.Wieser wrote:
Ant,

Someone told me it was a bad idea to everything on from a
power strip. Is this true?

(the below is an answer to the above, switching ON. Swiching off by it has
got its own, and already described, problems)

It depends. On what ? On how your computers power supplies respond to
it.

In the old days there was little done about the so-called "inrush" current
(the current a power supply could draw in the first few mSec its turned on -
which can be a number of times its normal one), and as such a few of such
power supplies together could reasily excess a classic fuses maximum spike
current, and blow out.

Nowerdays with all those (slow!) automatic fuses and (computer) power
supplies that are protected agains the effects of inrush current its not as
much as a problem anymore.

In my case I've got at least four PCs and a laser printer on the same power
strip (as well as several power warts, some feeding routers) which I'm using
every day (on in the morning, off in the evening), and have never had a
problem with it.

Regards,
Rudy Wieser

There are two ways to handle inrush.

NTC thermistor is the old way. Runs hot under normal operation,
starts out "cold" to work properly at the application of mains power.

The new way, is a side effect of the Active PFC controller. It's possible
as the power supply starts, to use the active PFC controller in a
series pass mode, for the purposes of limiting inrush. After a
suitable (short) interval, it becomes line interactive as it
attempts to make the voltage and current waveforms "in sync",
dropping the power factor to unity. (Older power supplies
have a power factor of 0.65, and current is not in phase
with voltage. In addition, the current waveform isn't
a sinusoid any more either and has a harmonic content.)

Someone mentioned power supplies with no switch on them.
My old Mac G4 works that way. It has no switch. After a short
delay after plugin, you hear a relay engage. Presumably the
interval between the two events, allows for a "slow charging"
of the main capacitor. So that, not even the relay gets
ripped to shreds by the inrush.

We had an inrush problem with a product at work, and the
responsible manager placed a *40amp* relay between the
mains and the (poorly designed third party) PSU. The 40amp
rating would be sufficient to "take inrush forever", and
there were never any reports of trouble with the relay
based design. It also made a satisfying "clunk" noise
at startup.

There was a rather large retail power supply reviewed at
Anandtech in the last year, and (apparently) the company
that designed it "forgot" to put any kind of inrush
solution in it. Anandtech reported that the breaker
in their breaker panel tripped, each time the supply
was powered up. Now, I want one of those, because I
haven't really tested the breakers in my panel for
some time.

Paul