| Dave Martindale 2008-02-13, 5:25 pm |
| A few weeks ago, I wrote about the UPM EM-100 energy meter. After
trying it, I ordered a P3 Kill-a-Watt model P4400, and it arrived this
morning.
At first glance, the Kill-a-Watt is the better tool for anything but
resistive loads. It does understand the difference between watts and
VA, and will display either along with power factor. The unloaded
isolation transformer that draws "20 W" according to the EM-100 uses 7 W
and 20 VA according to the Kill-a-Watt, with a power factor of 0.33.
(BTW, is this a typical power factor for transformer magnetizing
current? I don't have any other meters that measure PF, so I don't know
what's typical. A 0.33 PF is a 70 degree phase shift.)
In addition, the Kill-a-Watt manual explicitly says that the volts and
amps measurements are RMS. The UPM manual doesn't seem to say one way
or the other, so its V and A display are probably based on average
measurements, not RMS. I haven't yet tried them with a load that has a
weird current waveform to see if they differ.
On the other hand, the UPM meter has a number of user conveniences that
the basic Kill-a-Watt lacks. Here's a comparison:
P3 P4400 Kill-a-Watt
* RMS V and A measurements
* Calculates and displays W, VA, and PF
* Measures frequency in Hz
* Cumulative kWh and total time (plugged in)
* Line powered; a power failure loses all cumulative data
UPM EM-100
* V and A are probably not RMS
* Calculates VA only (and calls it W)
* no frequency
* Cumulative kWh and total *running* time (time current is not zero)
* Battery powered; no loss of cumulative data with power failure
* Records peak A and peak VA since last reset
* If you program your cost/kWh, it converts cumulative kWh to dollars
for you.
Having the cumulative time be running time for the device, rather than
total elapsed measurement time, means you can calculate average running
watts of a device like a refrigerator. Similarly, peak current might be
interesting.
There is also a P4460 model Kill-a-Watt that adds battery backup and
dollar cost functions, though it's somewhat more expensive. I haven't
tried it.
Dave
|