Measuring power consumption of low-power non-sinewave devices - how?

[tinnews]

Is there any cheap/easy way to measure the power consumption of mains powered equipment that consumes a few tens of watts and is almst certainly non-sinusoidal? In particular I want to measure the power consumption of a (hopefully) low power PC that should be consuming around 25 watts.

A cheap plug-in power meter (a Brennendal one) claims it's consuming about 50 watts but I'm not at all convinced it's telling me the truth.

I know 'proper' power meters are expensive and I don't really want to spend lots of money, nor do I really want a very accurate answer. If I can just confirm it's consuming something between 20 and 30 watts I'll be happy.

Can one buy kwh meters relatively cheaply? I'm quite happy to wire one up to a plug and socket and use that over several hours to see what's being consumed.

[brass monkey]

Surely you already have one in the meter locker?

[Dave Osborne]

[<me9]

What spec is that? my atom board one with one hard drive takes about 40 watts. I wouldn't expect much less other than for a netbook.

The Aldi/lidl ones I have seem quite accurate, but at low (

I compared my cheap ones with one and found them accurate. You should be able to pick a proper mechanical KWH meter for less than £20.

[pete]

You could always increase the power consumption to get it into the meter's linear range. For example, get some 60W bulbs (3 or 4) in parallel. Connect them up to the meter and measure the power consumption. Plug in the load from your low power device as well as the bulbs and see how much _extra_ consumption is taken.

[whisky-dave]

> I know 'proper' power meters are expensive and I don't really want to

I'm not sure how it calculates things but current and voltage are usually out of phase with computer power supplies so there's a few different reading that can be taken. The power in watts or kwh (over time) then there's the VA which is amps X volts which gives you power but this might not be the same as watts.

I just tried one of our small Dell PCs in the lab, when it's off but plugged in it's 10 watts and 125Watts when on, excluding monitor.

I find it hard to believe that a PC is 25-30w when on. My 24" iMac is about 110w changing the LCD screen from minium to maxium xhanged the power usage by nearly 20 watts. Doenloadign a file added 5watts as I guess ther HD was being accessed.

[Andrew Gabriel]

I found the Brennendal one to be very inaccurate. It was suggested to me that this happens if you don't fit batteries in it, but I never got around to seeing if that was the case.

Simply buy a different make. AFAIK, it's only the Brennendal which is very inaccurate. Get one which can give you the power, VA, and power-factor readings - it's likely to understand the concept of non-sinusoidal current loads.

I also still have one I made about 30 years ago, before the cheap plug-in ones appeared.

BTW, some time ago, I wrote a blog on the significance of power-factor, which you might find interesting, with some java applets you can play with changing various parameters and see what the effect is...

[Bob Eager]

Try a fanless VIA EPIA board, lower end, running off a CF card.

More like 15-20W.

[cynic]

Reconditioned kWh meters are available on ebay for around a tenner