Since Jim Wilkins mentioned Kill-A-Watt meter here some time ago it came on sale at Lee Valley and I got me one. I spent an interesting hour running around the house finding out what consumes electricity while plugged in and what does not. What surprised me were the computers.
My most recent purchase (Compaq Presario) does this:
Turned off 2.1W Booting 60W On but quiet 46W Asleep 5.8W Hibernating 1.7W
My old computer also draws about 3W while supposedly switched off.
So the questions are:
1) Why is the computer drawing any power at all when turned off?
2) Why is the power drawn less when hibernating then when switched off?
3) What happens to a computer which is turned off and the plug is pulled?
4) What happens with laptops? Do they draw power from their battery continuously even when turned off?
Why has this stupid Windows Live Mail program suddenly decided to type É instead of a question mark ÉÉÉ
In a discussion on alt.energy.homepower someone (Neon John?) wrote that they had tested a Kill-A-Watt against lab equipment and found it quite accurate, but like any digital instrument you can't trust that the accuracy is as good as the resolution. If the 0.01 Amp display resolution matches that of the A/D converter (doubtful) then the wattage at 120V would change in steps of 1.2W and the real accuracy is no better than +/- one step.
The circuit that commands the computer to power up when you press the momentary-contact power button has to draw some power itself. I believe they were designed for lowest cost rather than efficiency. The service manual for my Dell Dimension says to press the power button - after- unplugging the AC cord to discharge a large capacitor, before swapping parts.
See above. Your 2.1W and 1.7W readings don't really mean that much unless you have checked the calibration of that KAW somehow, like with a resistor load. Which I don't suggest. Even if you do the cheap methods of converting 120 - 240VAC to a small amount of low voltage DC can have high power factors.
The front-panel button won't turn it on until you restore AC power.
I have a main power strip on the side of this table that cuts off everything, slays the energy vampires. (OK, I'm a fan of Joss Whedon's work). The UPS and laptop charger are plugged into it, and sometimes a soldering iron and small heat gun. The two desktops are on separate strips plugged into the UPS. To use one I turn on the main strip, then the UPS and let it self-test, then the strip for that computer. Their monitors, USB drives, speakers, printers and keyboard lamp more than double the power demand.
This Latitude doesn't, the batteries stay at 100% when it's shut down or in hibernation. Standby does use some battery power. I recently had to replace the 2032 CMOS batteries in my 10-year-old Compaq laptops. Their main batteries are dead so I run them off AC and they wake up from hibernation with no battery installed, usually the instant I plug in the charger. Ergo they must not need battery power in hibernation. The only symptom of the dead CMOS battery was the 1980 date.
computers have soft switches. they are never off, just like a TV that has a remote control. they use a few watts just sitting there. The power supplies are on, in a low power mode.
A 400mW difference is pretty much nothing. A hibernating computer is as off as they can get while being unplugged. The tolerances of your chinese power supply before and after warming up could easily cause that difference.
A computer is a sleep state is more "on" than computer that is off, but far less active than one that is running and doing stuff.
nothing. it uses no power and does nothing.
yes. they draw power, but it's very very small- far less than the 1.7 to
2.1 watts a desktop might use when soft-off. The amount drawn is so small it doesn't even matter in real life.
Hibernation is basically off with the previous state of the computer written to disk, so no power needed to maintain that state. It's got a flag that tells the boot code that it was in hibernation, so it goes and loads that state up and theoretically, you're back where you were. Occasionally, the disk write goes haywire and it can't successfully read the hibernation file, so you end up like you would from a cold boot. Standby just powers the CPU and peripherals down to minimum power required, but maintains the memory contents as-is. Nothing written to disk. If you pull the plug, you've just lost all open files and changes. I hate standby. The trade-off is the almost immediate response when resuming, no need to read a hibernation file or run POST, it just powers up the CPU and disk and starts where it left off.
Laptops probably draw current when "off", they're checking that power button just like their big brothers. You could find out how much by connecting the wall wart power supply to the watt meter, get a not- connected figure, remove the battery from the laptop and then connect the power plug to it and get that figure. It isn't going to be much, might be less than what the wattmeter can read. If the battery is in, the charger is going to be trickling power into it to keep it charged.
Switching Mode power supplies are going to throw a monkey wrench in those accuracy tests. You have to test for that specifically.
You are also supplying power to the Real Time Clock chip so you don't run down the lithium battery on the motherboard, and the LAN card and it's "Wake On LAN" circuit used on certain industrial use machines - print spoolers and the like.
Some computers are also set up to power-up or wake-up when the phone on the modem rings, or power the modem and answer the line while the computer is still booting/waking. So it can sleep or hibernate and still be used as a fax machine, or be remotely polled at 2 AM in a business setting.
It's a switching-mode power supply, and you might be getting anomalous low or high readings at those low power levels. Do a little research on "Q Factor" where power panels with large numbers of computers burn up the neutral wires and the busbars in the panels.
The power supplies take little nibbles out of each cycle and the cumulative current on the neutral is higher than the power phase draw
- they don't balance out like normal phase-neutral loads, because each computer is taking it's nibble at a different moment in the cycle.
More prevalent on 120/208V 3-phase panels. They make special panels and special "High Q Factor" 480V to 120/208V transformers with double-sized neutral bussing to handle this. And you double-up the neutral conductors or bus-bars between them.
All well and good, but you'll be changing the CMOS battery more often. Better than the old days, when the battery was molded into the RTC chip, and you have to replace the whole thing.
Okay to kill the power totally to all the other peripherals, but leave the computer itself plugged in. Plus, you want it to be turned on overnight at least once a week - especially on Patch Tuesdays, when Mickeysoft tries to plug all the holes.
I just let it run, they seem to live a whole lot longer that way. So it costs a few pennies a day, big whoop.
They use a tiny bit too, but they have better power management.
Mon clavier did it only in Windows Live Mail - in the word processor it was "anglais". And stopped doing it once I closed down and restarted the WLM. Personally, I think WLM is un morceau de merde for many other reasons. Not as good as the old Outlook Express.
It took me a couple of weeks before I managed to stop the bloody thing waking me up when it felt like it, so no, thanks. It can jolly well patch when I tell it to. It's not just MS that wants to run when it pleases - others seem to want to do so, too.
The worst was the Window Media Centre: It was waking up every morning doing its thing and because the software was loaded wrong in the factory, there was no way to access it. After spending two hours on the phone with HP support I had to re-load the whole thing including all the programs I put on myself. I was not best pleased.
I guess you buy cheap crap you gotta expect these things. I would not have bought it if were not for the good reviews.
Because it is never truly off. There needs to be enough power on a couple of pins of the power supply so it can sense the pressing of the button to tell it to turn on. In the old days, the power switch was on the side or back of the computer, and truly disconnected it from the power line, so it was truly off. These days, the systems are set up to close down gracefully and then drop the power from the drives and the majority of the system board, so it needs a way to have the system board tell the power supply when it is truly ready to shut off.
Also -- it probably keeps power to the clock chip to keep it up to date, and keep the coin cell from discharging -- and to keep from losing the contents of the setup NVRAM, which is usually a CMOS RAM backed by the coin cell.
That one is a good trick. I don't know -- inefficiences somewhere, perhaps?
Once the plug is pulled, it can't sense you pressing the power button, so it can't turn on. If you shut the computer down, unplug the power cable, push the button once (which would normally turn it on) and then re-plug the power cable, it probably will not turn on -- unless there is something remembering that power switch press kept alive by the clock chip and NVRAM cell.
There is typically a coin cell buried in the computer to retain the clock chip and the NVRAM settings. If that goes dead, then it depends on the main battery. If that goes dead as well, the system will next wake up with no proper sense of time, and the NVRAM settings lost as well.
Also, the battery is used to power the switch which turns on the system, though most of the logic in the computer is CMOS instead of TTL, so it does not draw significant power when it is not changing state.
But the batteries used in laptops are known for self-discharge anyway -- the batteries draw enough power from themselves to dwarf what the NVRAM and switch sensor draw.
That's not the "good old days" - that's the "in-between nasties". In the good old days there was a big square CMos battery which could be replaced with a 4-pack of AA dry-cells. The self-powered RTC chip was, thankfully, a rather short term aberation - or was that an abortion.
Actually -- that chip was something which Sun kept for *way* too long. The clock chip actually is a TOD (Time Of Day) clock/calendar, not a Real-Time Clock which generates precisely spaced pulses for task switching in time sensitive applications like the CPU controlling a CNC machine tool. The two terms have gotten mixed up over the years.
The one Sun used (from Toshiba, IIRC) had 2K of CMOS RAM in addition to the clock and the potted-in cell on machines at least from the early SPARC-1 machines up through the Ultra-60 in my experience. The problem was that the RAM contained (in areas inaccessible for write by the firmwre) the HOSTID and the MAC address (raw ethernet address). And if you were using licensed software (some compilers and fancy backup systems were examples which I know of -- probably some database programs and CAD packages as well) depended on the HOSTID and the MAC address. It was possible to re-write those sections using a fcode program (FORTH) keyed into the NVRAM part, but it was a real pain. Each had a bar code label and you could order a replacement from Sun by barcode number. The rest of the 2K was used for the eeprom setup variables, and there was a keyboard sequence during boot to reset the variables to the factory defaults.
About the time of the SS-2, a company named Solbourne brought out a semi-clone which had the HOSTID and MAC address in a bipolar PROM, and the EEPROM and clock were backed by a coin cell which could be easily replaced. The bipolar PROM was mounted in a socket (as was the NVRAM/CLOCK chip on the SPARC systems), and if you had to replace a system board, it came without the bipolar prom -- you were supposed to swap it over to the new board so your licensed software would continue to operate.
Yes -- I had to replace a system board in a SS-5 which was running a licensed compiler, and so I swapped over the NVRAM/CLOCK chip when I changed. (It helps to have a collection of spare older systems sitting around when one dies. :-)
The Sun use of the clock/RAM chip started perhaps around 1980 or so, and continued until the Sun Blade 1000/Sun Blade 2000/Sun Fire 280R and later UltraSPARC systems, which put most of the information in a SEEEPROM (Serial Electrically Erasable EPROM) and the clock is powered by a coin cell in a holder on the system board (plus, of course, the power supply itself when the system is plugged in. :-)
So -- that Toshiba chip was in service far too long in Sun systems.