Twisted Mods Logo
  Site Map
Reviews 1
Reviews 2
Case Gallery
Submit News
About & Contact

Measuring Computer Power Consumption

Page: 1/1

HOWTO: Measure Your Computer?s Power Consumption

By Nick "Wixx" Pelis,

With so many people building custom, high-performance computers and using heavy, humongous monitors, one aspect of the computer often overlooked is its power consumption. Additionally, case mods such as water cooling and lighting systems have increased the demand for electricity even further. In this short HOWTO, we will take a look at how to measure your computer?s power consumption and the implications of the results.

*** Warning: AC current is very dangerous and can be lethal if not handled properly. Neither nor the author of this article are responsible for any loss or damage of property or life that may occur***

You will need the following pieces of equipment:

  • Standard computer AC Power Cord to cannibalize

  • Multimeter capable of measuring AC amperage

  • Pocket knife and/or wire strippers

  • Electrical tape

First, familiarize yourself with the layout of the electrical outlet:

Assemble together the parts you?ll be using:

For this exercise I?m using a standard AC computer power cord and a multimeter. If you don?t have a multimeter, spend the $20 and get one at Radio Shack; they are extremely useful tools for PC techs/case modders. I couldn?t find a pair of suitable wire strippers so I instead used my Leatherman.

Begin by stripping off the outer black insulation of the power cord across a 3 inch section. This will reveal three inner wires. Be very careful not to cut into one of the inner wires. To demonstrate that not every AC power cable is made equally, I stripped another one, and you can see the difference between the two in terms of shielding and wire color.

What we want to do now is determine which of the three inner wires the ?hot? wire is. Typically, this wire will be either white or blue. However, I?ve had some rather unpleasant experiences with AC electricity before and get rather paranoid, so I utilized my meter?s built in continuity tester, which causes the meter to beep if the two test probes are making electrical contact. To find the hot wire, connect one of the test probes to the hot contact on the plug, which is the rightmost contact if you are looking at the plug from the front (see picture). Then stick the other probe into each of the other wires until the meter beeps. Sure enough, in my case, the hot wire was the blue one.

Using your knife, scissors, or wire strippers, cut the hot wire but leave the other two intact.

Strip about one inch of insulation off both ends of the cut wire:

Wrap one of the exposed wire ends around one of the test probes. Because this is AC power, it doesn?t matter which test probe you wrap which end of the hot wire around.

Wrap the entire assembly together with electrical tape, taking care to ensure there is no exposed section of either the wire or the test probe. This is important, because you can kill yourself if you make contact with the hot wire while being grounded. This also ensures that the wire doesn?t slip off the end of the test probe while you are experimenting.

Repeat this procedure using the other test probe and the other end of the hot wire. You should have an assembly looking something like what I?ve got. Now it?s time to measure current.

Begin by setting up your meter to measure AC current. If yours has a separate port for measuring amperage (denoted by an ?A?), make sure to move the red test probe lead to that port, otherwise you could damage your meter. Then set your meter in AC current measuring mode, which is denoted by the ?A? with the tilde over it.

The setup: I hooked up my AC current measuring cord to one of my 15? monitors. Turn on the meter, plug the appropriate end into the monitor, and then plug the other end into the electrical socket.

The meter immediately tells us what we want to hear: the power consumption for the device it is plugged into. In this case, my monitor is using 0.8 amps when it?s turned on.

Turning off the monitor yields an interesting discovery:

Even when off, the monitor is still pulling a little bit of current. You will also see this if you plug your meter into a computer with an ATX power supply that is off, due to the standby current being sent to the motherboard.

What does this mean?

We can figure out how much it costs to run a device based on its power consumption and current electric rates. Out where I live, the cost of electricity is eight cents per kilowatt-hour as noted on the electric bill. We?ll use a formula that stipulates:

Watts = Voltage * Current (P=E*I)

There are 115 volts coming out of the wall electrical socket, and based off of previous measurements, we know the monitor consumes 0.8 amps when on. Therefore, the power consumption is:

Watts = 115 volts * 0.8 amps
Watts = 92

92 watts is actually 0.092 kilowatts (92 / 1000 = 0.092). Therefore, if it costs eight cents to use one kilowatt for one hour, then the cost to use 0.092 kilowatts for one hour is 0.736 cents (8 cents * 0.092 kilowatts = 0.736 cents). There?s 24 hours in one day, so running the monitor for one day continuously costs:

0.736 cents per hour * 24 hours = 17.664 cents

There?s seven days in one week, so running the monitor for a week continuously costs:

17.664 cents/day * 7 days = 123.648 cents, or $1.24

Finally, there?s 52 weeks in a year, so to leave the monitor on for one year would set you back:

123.648 cents/week * 52 weeks = 6,429.696 cents, or $64.30

Another use for this information is in planning a LAN party. Most household circuits are rated at 20 amps. To prevent the breaker from tripping, you should load up the circuit to no more than 80% of its rating. That means you have 16 amps per circuit. If a typical monitor draws 0.8 amps, and a typical computer draws 1.0 amps, that means you can safely fit about eight people on one circuit (approximately 2 amps per person * 8 people = 16 amps) without overloading it, assuming the circuit has no other loads on it.

Copyright © 2004,
Page was rendered in 0.001 secs.