First question is...
What is the configuration of your G5? Amount of RAM, number of hard drives, speed of hard drives, video card, any PCI cards.
Were any peripherals attached, that are powered by the Mac, such as scanners, hubs, docking stations, Etc.
Are you using a UPS or surge suppressor, and was that inline when you were doing your test.
Any of the above would change the amount of current drawn by your Mac. In most cases the problem is a parasitic load that is not accounted for in the testing.
I used a device you plug into to the power outlet and then you plug the computer into that and it will show you how many watts are used. You can also program the price pr. KWh and have it calculate the daily, monthly and yearly cost based on the average power consumtion of the unit plugged in.
The configuration of my Mac is in my signature - besides that I also have an AlchemyTV PCI card + a PCI USB card installed and a Edirol PC300 keyboard, A BT dongle, A logitech wireless mouse and a Canon S600 printer attached to the USB card.
I'll try to unplug all external units except the keyboard and see if it makes a difference.
The G5 below has these readings (240 volts):
Off - 33.6 W
Sleep - 40.8 W
Idle - 204 W
I'm glad of the warmth actually, as the boiler is on the blink:
Think what you're actually measuring is Amps, and converting it.
I havent read much about the on/off/idle wattage loads of the G5 but I'm an electrician and get asked alot "how much will all that landscape lighting cost me a month? " so I have to calculate it for them.
It's a pleasent surprise to read your finding is only 2000 KWh per year.
That works out to a meer $200 per year at our residential rate of .10 cents per KWh, or .06 cents per day.
Have you tried testing that wattage meter for accuracy with a known load of something like a 75 watt light bulb?
Your system idle:
200 watts X 24 hours = 4800 watts or 4.8 KWh
4.8 X 365 days = 1,752 KWh
1,752 KWh is cheap.
I know your question was wondering why the sleep and off loads are higher than expected, but $200 a year is a bargain.
I'm curious, so let us know what you find after testing the meter with a light bulb.
I've found most our customers have a misconception that they're computers are causing high electrical bills.
Apparently some of this idea comes from many years back when computers were looked at as a novelty and voodoo because most people didnt understand what they are.
In reality, most the increased cost caused by computers is the increased cooling loads generated by the box and monitor.
That's why my CRT only gets turned on when I need it. The office will climb over 80 degrees F in a couple hours with the room closed if I dont.
Look here for the source of your high electric bill and anywhere else that could be attributing to it like a old refrigerator or electric ovens and dryers.
edit: something doesnt appear to be right with the math too.
should be approx .60 cents per day
You might find 2000KWh pr year to be a low number, but our households total usage is 5200 KWh pr. year so it's more than 30% of the total usage.
According to statistics our electrical usage should be closer to 4200KWh / year so 2000KWh IS a lot (maybe not in the US, but for a Danish household it's quite high).
But no matter what it doesn't change the fact that according to Apple the power consumption SHOULD be much lower:
So even though you think $200 / year is cheap (the price in Denmark is closer to $500 for that amount of electricity - NOT CHEAP) that's beside the point, because the machine IS using much more power than it should, and as you can see from a post above, I'm not the only one with these figures.
The questions is - Is Apple lying on the page referenced above, or is there something wrong with our machines?
Reference measurements (240volt) - Fluke digital multimeter
60watt tungsten lamp - 55.2 W
100watt tungsten lamp - 96 W
2 x 500watt tungsten/quartz/iodine/halogen lamp
A - 458.4 W
B - 482.4 W
1050watt(nom.) toaster - 1012.8 W
laserprinter 480watt (nom.)
idle - 7.3 W
printing(avg.) - 384 W
peak load - 592.8 W
So that's why APC advise not to connect laserprinters to the UPS...
And... the Mac via APC Smart-UPS 750VA (240volt):
Off - 100.8 W
Sleep - 108 W
Idle - 256.8 W
compared to Mac only:
Off - 33.6 W
Sleep - 40.8 W
Idle - 204 W
This morning, I plugged one of those gadgets into my APC RS1000 UPS, which handles the SP PowerMac G5, the Sony CRT monitor, and a 14 inch iBook. With iBook, G5, and monitor on, but not doing anything CPU intensive, the total wattage was around 250-260. (I have the CPU performance setting at "highest".)
Shutting off the computers, but with the monitor still on brought it down to 85 watts or so. After a minute or so, the monitor went to power saving mode, and the total consumption dropped to 23 watts. I guess that's the residual load with a trickle charge into the APC and into the iBook battery. If the batteries were recharging I'd expect the usage would be higher.
The SP G5 does seem to use a lot less current than my older Athlon 2700+ box, which can keep the temperature up 5-10 degrees even on a cold windy day.
I've been noticing this as well. My Dual 2ghz G5 is pulling ~170W sitting idle. This is as opposed to my Sawtooth G4 1.4ghz (w/ upgrade) which runs at ~50watts idle, and my dual 550mhz PIII server (currently w/o drives) which is also pulling ~50 watts idle. Makes me a bit guilty about using my G5 so much.
While sleeping the G5 does drop to 10 watts, so that isn't as bad. Looking forward to an Intel Merom/Pentium based future....