(In)Accuracy of a Kill-A-Watt for measuring low power usage

[Winston]

Apparently Kill-A-Watt and other similar devices to measure electricity use are not accurate at low wattages.

Searching for info on how much power "wall wart" transformers use on standby, I ran across this note from the Lawrence Berkeley National Laboratory:
http://standby.lbl.gov/measure.html

"Measuring Standby
Few products indicate their standby power use, so you need a meter to measure it. Unfortunately, few meters have sufficient resolution to measure standby power accurately. These cost over $500.
A few meters are now available for less than $100. Unfortunately their accuracy below a few watts is poor. These are nevertheless indicative of standby power. You can improve accuracy by adding a standard load (such as an incandescent light bulb) into the circuit; then measure the difference in power consumption."

I assume this applies to a Kill-A-Watt. So if you want to measure a small transformer's standby energy use, put a known wattage device like a light bulb on the same circuit, then measure the difference between what the bulb uses and what the Kill-A-Watt shows.


The LBL site also has some interesting charts on standby electrical use by different devices:
http://standby.lbl.gov/summary-chart.html


Good luck.

- Winston

[space-time]

Mine shows powers as low as 3W. I have a Roomba that used 28W when charging and it drops to 3W when done. I also checked with a MAHA battery charger, which takes about 10W when charging 4 AA cells at 1000 mA (each cell needs 1.5W and with power loss in electronics, etc) that makes sense and it drops to 0 or 1 W when done.

This is good enough for me for a $20 device.

Just a few days ago I measured the power consumption of my subwoofer when sound is off and it is about 11W, so I have now I power strip with Master and Slave outlets that I plan to setup this weekend to cut off power to subwoofer when Receiver is OFF.

[Winston]

I'd forgotten about the subwoofer on our TV system. One more thing to look at on power consumption. It's wireless, so would have to stay in a standby mode when it's on. I think it has a power switch.

The issue with all these devices is how to put them on switches with reasonable usability. Most rooms in our house have A.C. jacks where one outlet is switched, but we use that for the lights - it would cause confusion if I put the subwoofer or other devices on that too.


- W

[freeradical]

I've used clip on ammeters attached to digital voltmeters to measure small amounts of current. Small as in milliamps.

[OWC Jamie]

Yeah, the kill-a-watt measures better averaged over time.

I've noticed it with small load devices where I've left everything plugged in for a month. The thirty day cumulative wouldn't math out the same as a momentary or short term measurement. I always figured the long term average was more accurate.

[Filliam H. Muffman]

It's a typical problem. Nothing that is "inexpensive" will efficiently measure 1500 Watts and 1.5 Watts. A clip on Amp meter that will accurately and precisely measure 200 mW and up to 1500 Watts will be $50 and up.

[lost in space]

Winston wrote:


I assume this applies to a Kill-A-Watt. So if you want to measure a small transformer's standby energy use, put a known wattage device like a light bulb on the same circuit, then measure the difference between what the bulb uses and what the Kill-A-Watt shows.


The LBL site also has some interesting charts on standby electrical use by different devices:
http://standby.lbl.gov/summary-chart.html


Good luck.

- Winston

That method could be useful, but only if the error was constant over a range of low wattages. More accurate would be to check wattage with a known higher load and the device together, then compare that with just the known higher load. The difference between the two would be wattage for the device. Possibly not dead accurate, but I'd have more faith in this method. My .02.

[Winston]

lost in space wrote:
[quote=Winston]


I assume this applies to a Kill-A-Watt. So if you want to measure a small transformer's standby energy use, put a known wattage device like a light bulb on the same circuit, then measure the difference between what the bulb uses and what the Kill-A-Watt shows.


The LBL site also has some interesting charts on standby electrical use by different devices:
http://standby.lbl.gov/summary-chart.html


Good luck.

- Winston

That method could be useful, but only if the error was constant over a range of low wattages. More accurate would be to check wattage with a known higher load and the device together, then compare that with just the known higher load. The difference between the two would be wattage for the device. Possibly not dead accurate, but I'd have more faith in this method. My .02.

That makes sense. I assume that, for example, there are variances in what a "100 W" bulb draws. Testing a specific bulb, and then that bulb with the low-wattage device would give you the difference.

- W

[AllGold]

I could be wrong, but I think the VA (Volt-Amp) setting gives you a little more "resolution" than the Watts setting.

But I suppose you're better off with a cumulative reading over a longer term.

[Racer X]

I have always run the Kill a Watt setup for a day or two, and divide the power consumption by the time to get average consumption.

You could easily check a freezer and read it while the compressor is running, or not.