John O wrote:
running normally, but idle: 145W to 160W, but usually 150W
all hard disks spun down, display blanked: 124W
Stand By mode: 60W
Hibernate mode: 2-3W
totally "off": 2-3W
after I flip power supply's rocker switch to "0": 0W

That's a cool set of measurements, and tells us a lot. The only factor that
isn't accounted for here is the monitor power.

DPMS monitors can basically be in one of five states. The first
four are specified by DPMS:

1. on (there's a picture)
2. stand-by (electron gun not firing electrons at phosphors,
but the power supply is on)
3. suspend ((gun's?) power supply is off)
4. "off" (but ready to wake up).

And the fifth:

5. unplugged or turned off by a mechanical switch

In theory, these could range all over the map, but in practice
there is a huge gap between stand-by and suspend. In other words,
"on" and "stand-by" both use 90%-100% of the power, and "suspend",
"off", and unplugged use 10% or less of the maximum power.

Anyway, some tests. First, tests of an old (mid-1990's vintage)
MAG Innovision 21" CRT monitor:

on, regular desktop color, 1280x1024: 120W
on, but with the screen mostly full of white: 144W
on, but with the screen mostly full of black: 111W

on, regular desktop color, 800x600: 100W
on, 800x600, white screen: 125W
on, 800x600, black screen: 92W

after windows turns it "off" due to no user input:
drops to 76W, stabilizes there for ~5s,
then drops to 37W
(and monitor indicates "POWER SAVING SUSPEND"
on its offscreen LCD display)

after putting windows into "standby" mode:
same 76W, then 37W behavior as above

after putting windows into "hibernate" mode:
drops to 76W, but then down to 27W;
(still says "POWER SAVING SUSPEND")

after shutting windows down completely:
drops to 76W, then 27W

after I flip the rocker switch on the back of the computer:
still 27W

after I punch the mechanical power button on the monitor:
0W (finally)

Conclusions from that:

1. All you people who tell us hacker people we are weirdos for
having green or amber text on a black background? Well, now
you know that by using that so-called weird color combination,
we are actually saving the environment, so THERE!

2. The resolution makes a difference. This is a little surprising
since it's still the same area that's being lit up. Apparently
making the beam scan faster (even though it's the same strength?)
makes a difference.

3. This monitor's power-saving support SUCKS. I did a similar
test on a newer CRT monitor once, and it drops down to like 3-5W
in situations where the above monitor only drops to 27W. So,
this monitor is probably not representative of CRTs in real life.
It probably doesn't implement DPMS "off" mode and substitutes
"suspend" instead. (But, I only paid $25 for the thing, so
I guess I can't complain *too* much.)

4. Since I leave the computer this monitor is attached to off
most of the time, the difference between 27W and 0W is pretty
huge for me. I pay about $0.10/kW-hr, so this 27W difference
costs me something like $2.00 a month!

5. My Kill-a-Watt meter, which was I think $25 or so, just paid
for itself by detecting that 27W that I didn't know the monitor
was sucking down while it's not doing anything, assuming I
can get in the habit of hitting the power switch when I should.

And now, some more tests. This time on a new Samsung SyncMaster
213T hooked up with DVI. This monitor is 21" nominal size, but its
actual usable display area is actually more than 1" larger than the
21" CRT mentioned above.

on, regular desktop color, 1600x1200: 40W
on, with screen mostly white: 40W
on, with screen mostly black: 40W

800x600: 40W

after mac turns it "off" due to no user input: 1W

if i turn if "off" with the monitor's front panel button: 1W

Conclusions from that:

1. This LCD is using about the same amount of power at its peak
usage as the CRT is when it's in "POWER SAVING SUSPEND" mode.

2. If you want to save power and you use the computer a lot,
the difference between "on" for the LCD and CRT is *huge*.
At my $0.10/kW-hr rate (and having the monitor on about 12
hours a day, since I use it for work and personal stuff both),
I probably save $3 a month just by going to LCD. Sure, it's
going to take me a while to recoup my $600 investment, but
it's nice to know I am at least saving some money.

Um, I think that should be enough info to cover everything. If anyone
wants to know how much my refrigerator uses let me know... ;-)

- Logan