Measuring PC power usage?

You can buy (at least in the UK) plug in power meters.

Definately not.

Watts (power) is already a rate of energy use (Joules/second). Hence watts/hour doesn't make any sense (unless you are referring to rate of change of power usage, which you aren't).

If you said what spec these items are, someone who has measured similar spec items can probably give you a ball-park figure.

search for "kill a watt" meter

that's the max rating of the supply.

if your load is 1W and operated for one hour that would be 1 W-H (watt-hour) if the 1 watt load operated for 1000 hours is then 1 kW-H (killowatt-hour) if a 1000 watt load operates for one hour it's 1 kW-H

The easy way is to buy a plug-in power meter. In the UK you can get them for

Cool, I googled for them and found one for £16.50 - that'll do nicely!

BTW, don't buy the Brennenstuhl one -- it's horribly inaccurate for loads like PC power supplies. I also have a no-name Taiwan one from Maplin, and that's OK. (Maplin used to sell the Brennenstuhl one, but stopped some time ago.)

it's a PM230, but not sure of make...?

Oh dear, that's the model number of the Brennenstuhl one. Test it with a compact fluorescent lamp, and if it reads 2 - 3 times the power rating of the lamp, then it's as crap as my one.

I'm not entirely sure what's wrong with it. I did play with it for a while when I first got it, and my impression is that it measures some type of average current, some type of average voltage, and the phase shift, and works out power as I x V x cosØ. It is failing to realise that low power factor can be caused by things other than phase shift, as is indeed the case with many PC power supplies (and integral electronic ballast compact fluorescent lamps), which have low power factor, but cosØ = 1 (almost).

You can find true RMS clamp on multi-meter in the used market for $20-30US. That is also going to be good for lots of other things.

If it is a name brand the manufacturer will usually swap it out for a rebuilt one for about $20-30 too so buying a broken one in a garage sale for a buck is a bargain. I got one out of the trash and sent in in.

Um, I bought the Maplin Brennenstuhl. :-(

ISTR Maplins sold it in two versions: one with the mains plug above the buttons and powered by two button cells. The other with the mains plug below the button and powered by AA cells.

I have an 18W fluorescent lamp and my PM230 reads just under 16W. Don't know exactly what's in a fluorescent lamp but I would expect the transformer and other things to need power in addition to the bulb so I would expect a reading of 18+ watts.

I notice that the cells must be in the unit or the readings will be very different.

I had thought the button cells on the unit merely kept the clock and pricing data and that it was ok to power the display from just plugging the unit into the mains. However I find there is a big difference in readings (about x2) with and without the cells.

If the CFL does not have proper Hg pressure control then the power will drop when the system is fully warmed up.

On Tue, 27 Jun 2006 19:51:00 +0100, Jon D Gave us:

A lamp does not have to be fully driven to light. It may be being pumped at 12 Watts with 3 Watts of losses. The power module would get a bit warm at that level, however.

Buy a meter. They are around $20-$30 at the local hardware store.

TTYL

I have a similar brenn* one. Also a similar model of a different make. Both misread by 100% or so on 2 of the 4 PC power supplies here. They both read correclty on all resistive, and low-power SMPSs - like CFLs.

Andrew,

Can you say how you established that the Maplin one was accurate at low power.?

John Stott.

Andrew Gabriel wrote: