Does frequent rebooting harm a computer?

[kony]

No I don't think so, they explicitly discuss all stages of power
management, what it does and how much power it saves. The mention of a
blank screen mainly comes up because in the absence of power management on
the computer, the iiyama can be set to detect a blank screen and
automatically go into standby mode after 3 minutes.

I don't think the term blank screen saver is misleading, since in Windows
at least turning on monitor power management is distinct from simply
choosing a screen saver, you can do either or both, and the "blank screen
saver" is just one of the screen savers you can pick from the list.

True, it does depend on whether you're just projecting a
black picture from a windows "screensaver" menu choice or no
imaging signal, a blank. The latter is supported by modern
monitors and allows lower monitor power state. I was too
ambiguous previously when I wrote "blanked black", I
should've omitted the word "black".

 

[kony]

The book explicitly says a blank screen saves a maximum of 20% power.

I feel we have to differentiate between "blank" meaning a
black picture signal is being sent (which a user could call
"blank") and "blank" meaning the monitor is displaying
*nothing*.

It seems that manual makes no mention of whether a screen
saver is being considered as only a MS windows setting or
non-display of a static image, including (excluding) a
static black image?

 

[John Jordan]

Eh? Where are you measuring?

That's with the monitor switched off with its own power switch, not at
the wall. With the cable pulled out, it does read zero. I guess the
problem is that some transformer circuitry is still connected.

This isn't unusual - I checked a couple of other CRTs here and a TFT,
and even the TFT pulled 10W when switched off. Similarly PC PSUs consume
10-20W with the machine powered off, just from transforming the +5VSB.

 

[kony]

It's the Amps that melt the wires. Ask the generating board why they
transfer leccy by 400,000 Volts.

No, having 50A on wires designed to carry 50A is no more
prone to melt wires than having 1A on wires designed for 1A.

In other words, it is irrelevant how many amps it is.
It is also irrelevant how how the CPU or anything else gets,
because relatively they will not, ever, get hot enough to
start a fire. An electrical malfunction could but that is
an entirely separate issue and not precipitated by the heat
of a CPU or amps expected on the power leads. Most often it
would be the result of a higher voltage arc, the kind of
issue present in many devices operating off of AC that
aren't even remotely like a PC.

 

[Jaimie Vandenbergh]

Nope, the issue was one of heat and failure being addressed
by powering off the equipment,

Quite right; I hopped subthreads without noticing. I still say
mxsmanic was being pointlessly picky.

You are also arbitrarily claiming things like "200W".

Hardly arbitrary. I was looking at the machine that I'm typing this
on.

Jaimie (Asus PCH-DL with two 3GHz Xeons, 533fsb, 1.5gig RAM,
2x.25Tb disks, nV6600GT)

 

[Jaimie Vandenbergh]

That's with the monitor switched off with its own power switch, not at
the wall. With the cable pulled out, it does read zero. I guess the
problem is that some transformer circuitry is still connected.

They just don't make off-switches the way they used to!

This isn't unusual - I checked a couple of other CRTs here and a TFT,
and even the TFT pulled 10W when switched off. Similarly PC PSUs consume
10-20W with the machine powered off, just from transforming the +5VSB.

I'm used to TFT's having external soap-on-a-rope style transformers,
so that wouldn't have surprised me. And modern PC's have all that
standby and "Wake on..." functionality that needs +5v active, as you
say. I expect that a power switch on an ATX PSU would be a true air
gap.

I'm surprised about CRTs not having mains breaker power switches, I'll
have to change my habits. Um.

Cheers - Jaimie

 

[kony]

Quite right; I hopped subthreads without noticing. I still say
mxsmanic was being pointlessly picky.

Perhaps picky relative to what you were thinking, but based
upon what was written it did seem relevant enough to
consider other appliances/etc as well as PCs.

Hardly arbitrary. I was looking at the machine that I'm typing this
on.

Ok, that's quite relevant for you, but arbitrary relative to
any other user's dissimilar system.

 

[kony]

I'm surprised about CRTs not having mains breaker power switches, I'll
have to change my habits. Um.

Some do, but in the rear- usually the front switch is logic
switch and has been, on average monitors, for roughly 4
years. I'd seen spec sheets that listed less than 10-15W
though, IIRC 4W was common.

By having a logic on instead of manual, it makes it easier
(though perhaps not essential) to implement power-off
command remotely from the PC.

 

[Mxsmanic]

It's the Amps that melt the wires. Ask the generating board why they
transfer leccy by 400,000 Volts.

Because using voltage to transfer watts requires smaller cables than
using current to transfer watts. The practice is not without
disadvantages, though. Five volts can't force enough current through
your body to be a threat to your safety, but 400 kilovolts definitely
can.

In any case, the total energy is limited to the product of voltage and
current, so high amperage at low voltage is no worse than low amperage
at high voltage overall.

More importantly, computers are no more of a risk than any other
appliance left plugged in.

 

[Mxsmanic]

You're being needlessly snarky. In order: near zero power, near zero
unless in use, near zero unless in use, and you don't _want_ to turn
your fridge off to save money, because it'll cost to replace the
contents.

They all use power continuously; if something goes wrong, they might
overheat, just as a computer might. Some of them contain motors, just
as a computer does. Refrigerators contain substantial motors, much
heftier than a computer does.

In other words, leaving a computer on presents no special risk
compared to leaving anything else on.

 

[Mxsmanic]

Terminology problem, I think. A "blank" screen saver at worst is just
a black screen, with all the monitor circuitry running at full blast
but not painting electrons. That's only going to save power at the
electron guns, so I imagine 110w would be about right.

It may save some power at the electron guns, but not much, since the
main source of wear and tear is the heating of the filament, which
continues even if the electron beam is completely shut off.

 

[Mxsmanic]

My wattmeter agrees with you. For a 19" Dell-branded Trinitron:

High brightness, 100% white screen 129W
Low brightness, 100% white screen 124W
Low brightness, "blank screen" screensaver 98W
Power saving 24W
Monitor switched off 17W

It's interesting to note that my flat-panel display of the same size
consumes a maximum of 58 watts when running full blast and a minimum
of just 3 watts in power-saving mode.

 

[David Maynard]

Jaimie Vandenbergh writes:




They all use power continuously; if something goes wrong, they might
overheat, just as a computer might. Some of them contain motors, just
as a computer does. Refrigerators contain substantial motors, much
heftier than a computer does.

In other words, leaving a computer on presents no special risk
compared to leaving anything else on.

Your analysis is overly simplistic because you do not take into
consideration the susceptibility of the device, it's complexity, it's
robustness, the reliability of the safety mechanisms, fault conditions,
cascade effects (such as location) and the resultant risk vs the benefit of
unattended operation. Not to mention the intended design (e.g. it is
expected that refrigerators will be run continuously for decades on end and
so they're designed to)

For example, a simple thermal cutout, such as on an electric motor, is
infinitely more reliable (as is the motor) than an electronic
thermistor/diode monitoring system running on a PIC in a motherboard nor
does it suffer from the human factor of some yahoo turning it off because
the alarm was annoying nor do they ever go 'insane' from a power glitch or
chase their own tails in a program loop. Nor do they ever need a 'BIOS'
flash to fix bugs, they're immune to 'viruses', never get a backdoor
trojan, present a BSOD, pause for a core dump, or go off-line for a reboot,
mysteriously clearing parameters in the process, then hang in a brain dead
state displaying "disk not found."

In short, they're not comparable simply because they use electricity in
some manner.

 

[DevilsPGD]

In message <[email protected]> kony

Actually it's often more cost effective to use a lower level
of lighting and NOT turn them off in habitated areas when
using fluorescents, and to not use compact fluorescents.

Not to be argumentative, but no it's not.

Unless you're using less then 13W, it's more efficient to use a compact
fluorescent. Leaving it on vs toggling is more debatable, depending on
the startup power vs the duration you'll leave your lights off.

 

[DevilsPGD]

In message <[email protected]> Mxsmanic

It's total watts that count, and in any case, there are many other
appliances that use at least as much power. Many appliances contain
electric motors that will consume a great deal of power if they seize
up. Other appliances can become very hot if something goes wrong
(electric water heaters, particularly instant heaters). There are all
sorts of things that pose a hazard if you're worried about leaving
things connected to power. Computers are no more hazardous than
anything else.

On average I'd say a lot less -- There are only a small number of things
that can go wrong that wouldn't draw too much current and short, or just
burn out a circuit pathway.

In either event, most quality power supplies will shut down safely and
chances are nothing much more then a burned out system will occur.

 

[DevilsPGD]

In message <[email protected]> Jaimie

I'm used to TFT's having external soap-on-a-rope style transformers,
so that wouldn't have surprised me.

Soap-on-a-rope? Meaning an external "brick" type?

All of the ones I've ever owned use the same power connector as a PC,
and have their transformers built-in. My current monitor (Dell 2005FPW)
has a built-in USB hub (always-on) and speakers (which can be configured
to turn off when the monitor is off or not), so it always draws some
power.

I'd much rather the USB hub go off when the monitor is off (I just use
it for power, I have a backlit mousepad and a USB snakelight that shines
on Dilbert that are connected through it's USB ports), but I don't have
that choice.

It wouldn't surprise me if the monitor draws as much power as my system
when they're both off.

And modern PC's have all that
standby and "Wake on..." functionality that needs +5v active, as you
say. I expect that a power switch on an ATX PSU would be a true air
gap.

Most of the time, yes. I've seen at least one in my travels that simply
interrupted +5V to the motherboard, so it was no more off then by
shutting down the PC (although it did manage to prevent the system from
suddenly coming to life if you triggered a wake up event)

 

[kony]

In message <[email protected]> kony


Not to be argumentative, but no it's not.

Unless you're using less then 13W, it's more efficient to use a compact
fluorescent.

Compact fluorescents are less efficient than full sized
fluorescents and "one" 13W CF is not enough to light up a
room to the level most anyone not accustomed to living in a
cave, will want. You'd have to contrast a larger number of
CF to smaller # of full sized fluorescents.

Leaving it on vs toggling is more debatable, depending on
the startup power vs the duration you'll leave your lights off.

True, plus the actual wear from repeatedly cycling, the cost
of replacement ballast (or whole CF "bulb"), as well as the
energy to produce, distribute and sell the product.

 

[Conor]

This isn't unusual - I checked a couple of other CRTs here and a TFT,
and even the TFT pulled 10W when switched off. Similarly PC PSUs consume
10-20W with the machine powered off, just from transforming the +5VSB.

Indeed. An article on an ecological website some months ago was
commenting on the amount of electricity consumed by common household
devices turned off which is now nothing more than a standby feature.

 

[Mxsmanic]

True, plus the actual wear from repeatedly cycling, the cost
of replacement ballast (or whole CF "bulb"), as well as the
energy to produce, distribute and sell the product.

My desk lamp has a compact fluorescent lamp inside. I leave it on
continuously. It consumes nine watts. The first lamp lasted for nine
years (nearly 80,000 hours). I'm on the second lamp now. The lamp
cost about $20 to buy, and it costs about 40 cents a month to operate
(at 24 hours per day).

 

[kony]

My desk lamp has a compact fluorescent lamp inside. I leave it on
continuously. It consumes nine watts. The first lamp lasted for nine
years (nearly 80,000 hours). I'm on the second lamp now. The lamp
cost about $20 to buy, and it costs about 40 cents a month to operate
(at 24 hours per day).

That's pretty good, though it may have had a bit to do with
the (low) wattage. I recall one CF I bought that was
recalled due to overheating/hazzard. It's replacement still
gets pretty hot but (without tearing it open I can only
speculate) appears to have primarily changed by using a
plated hole PCB.