Message-ID: <01FD6EC775C6D4119CDF0090273F74A4022028@emwatent02.meters.com.au> From: "da Silva, Joe" To: "'opendos AT delorie DOT com'" Subject: RE: Power consumption #2 (was Power measurement) Date: Wed, 11 Apr 2001 14:11:25 +1000 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2448.0) Content-Type: text/plain Reply-To: opendos AT delorie DOT com Yes Alain, "reactive power" does have _some_ effect, as I already mentioned (see "resistive losses", etc.). However, it is still just a second-order effect, so 1 VAR has _much_less_ effect than 1 W ... Thinking about it, I don't think it is accurate to say that PC supplies are "capacitive" ... much of their "reactive power" is due to harmonic currents, rather than current phase-shifting ... Also : - A capacitive load has peak current around the zero-crossing region of the input voltage. PC supplies have peak current around the peak region of the input voltage. - A capacitive load returns it's energy to the "power grid" during half the voltage cycle. A PC supply never returns any energy to the "grid". So, in other words, ordinary PC supplies are "nasty" and provide little compensation for other inductive loads on the "power grid". There are now supplies available with "power factor correction", which have "nice" (resistive) characteristics. However, I don't know if they are used by the PC manufacturers (they cost more). Joe. > -----Original Message----- > From: Alain [SMTP:alainm AT pobox DOT com] > Sent: Tuesday, 10 April 2001 7:47 > To: opendos AT delorie DOT com > Subject: Re: Power measurement (was [off-topic] shutting down) > > >Wow! For an O/T, this stuff about SI/Imperial/American terminology > >and measurements sure generated a lot of messages! > > > >Anyway, back to the original topic ... > > > >You cannot determine the power consumption of PC's and monitors > >by just measuring the current - even if you measure this in RMS! Of > >course, if the current measurement isn't RMS, then that's even more > >useless! > > > >If you measure the RMS current and multiply by the (RMS) voltage, > >what you have is the "apparent power" (VA), not the "real power" > >(W). That's the reason why the "power" figures quoted below are > >higher than expected. Now, "apparent power" (VA) is the _vector_ > >sum of "real power" (W) and "reactive power" (VAR). > > > >So, what is "reactive power" (VAR), and what impact does it have? > >Well, it's the out-of-phase current ... it is current that recirculates > in > >the power system but, apart from resistive losses in the wiring and > >generation system (ie. second-order effects), does not require any > >input (eg. coal) to the system. If we used superconductors for all > >this stuff, then even these resistive losses would disappear! > > > >So, the only way to correctly measure the power consumption of > >PC equipment, is to either use a proper power meter (hard to find) > >or use an energy meter (eg. the one at the front or side of your > >house) and see how much energy it accumulates in one hour (if > >necessary, run the test for several hours, then divide the energy > >accumulated, by the number of hours). > > It is true that the "reactive power" is not real power, but only at the > consumer end of the line. From the other side (where it is generated), > this makes more energy to be lost in the distribution wires. What > happens then is that if you have too much "reactive power" you will > be billed extra (something like a fine). On the other hand, PC supplies > have a shift to the capacitive side which can somehow compensate > for the inductive power used in most industries and contribute to > fewer losses. I believe only a macroscopic view of the problem could > get close to reality and know of no study about this. > > One good way of measuring power consumption is by the bill you > receive (a philiphical point of view), another one is the ecological > impact. I believe the later is better and by this standard we should > turn everything off when not in use ;-) > > Alain >