rsd@sei.cmu.edu (Richard S D'Ippolito) (01/03/91)
In article <5402@rsiatl.Dixie.Com> larry@rsiatl.UUCP (Larry Kahhan) writes: >>rsd@sei.cmu.edu (Richard S D'Ippolito) writes: >> >>>... Again, there is NO SUCH THING as "true RMS"; it's just RMS! ... >> > >Not so! Early AC voltmeters used a technique to measure the RMS values >of sinusoidal waveforms by a simple calibration technique. Uh, Larry, how old are you??! At best, what you meant to say is mis-stated by confusing measuring technique with calibration technique, but I catch your drift. "Early" AC voltmeters of the iron-vane repulsion or thermopile type co-existed with and in many cases preceeded some of those average-responding D'Arsonval movements so common until the advent of digital display meters. These were in common use in various industries, including the electric utilities and standards labs. (For those who may not know, the "simple calibration technique" is insuring that the calibration source is a pure sinusoid and adjusting the output scaling so that the rms-equivalent is displayed.) The lower cost of the "newer" meters made them available to a wider, but not necessarily more educated clientele, who weren't aware that the analog (and now digital) meters would not read the correct RMS of a non-sinusoid, now more important as the resolution increased (mostly far in excess of even the short-term accuracy) by using digital displays. >When AC >voltmeters came along which could determine the RMS value of an >arbitrary waveform (using any one of a number of techniques) these >meters were called "true RMS" to distinguish them from ealier RMS >meters. Nope, the goofy phrase was coined by a combination of marketers and at-best apathetic engineers who didn't insist on "RMS-responding" (doesn't sound as catchy, does it?), and now we're stuck with a re-education task, always harder the original educating would have been. Your KWHr meter has always been RMS-responding. >However, there is only one real definition of RMS, as >Richard points out. > >Larry Kahhan - NRA, NRA-ILA, CSG, GSSA , & GOA Other goofy marketing anomalies (or should I just be honest and call them the mistake they are?) are the 3-1/2 (or whatever) -digit tag on a meter whose full scale reading is 1999 counts (anybody know logarithms?), and the meter with "0.1%" accuracy, enthusiastically promising a reading no more than a factor of 1000 off! Rich
news@pasteur.Berkeley.EDU (Six o'clock News) (01/03/91)
>Your KWHr meter has always been RMS-responding. >[...] >Rich RMS is not and should not be used in 'true' power meters. They read just plain average power ( I will use '<P>' for this). The only reason people use RMS to begin with is to be able to calculate <P> from V^2/R by using V=RMS(V(t)). From: charless@cory.Berkeley.EDU (Charles R. Sullivan) Path: cory.Berkeley.EDU!charless Perhaps this example will help. To spare you the trigonometry, I will use a square wave voltage, 50% duty cycle, between 0 and 100V. RMS(V(t)) = sqrt( (100^2 + 0)/2 ) = 70.7 Volts. With a 10 ohm load (resistive), I(t) is a sqare wave between 10A and 0, so <P(t)> = <V(t)*I(t)> = (100*10 + 0)/2 = 500W To calculate <P> = V^2/R, we use RMS(V(t)) = sqrt( (100^2 + 0)/2 ) = 70.7 Volts, so <P> = 5000/10 = 500W, as before. We needed the RMS trick, because the straight average voltage, 50V would not give the right result; 50^2/10 = 250 W. However, doing RMS(P(t)) yields a wierd, not-useful result: RMS(P(t)) = sqrt( (1000^2 + 0)/2 ) = 707W. I certainly hope the power company doesn't use that to bill me. (That would mean, for example, that using a 1000W appliance for 1/2 hour would be billed the same as a 707W appliance for 1 hour!) ~ Perhaps the reason a 'true RMS power meter' is sometimes specified is that if you made a power measurement by measuring voltage and current, with separate meters, and then multiplied the measurements together *after* the meters averaged each independently, you would prefer meters that used 'true RMS' measurements of the voltage and current. This would result in correct average power in the case of a resistive load. However, it would not give correct results for a current waveform that was not identical in shape and phase to the voltage wavform. Hence the need for power meters, which multiply the instantaneous voltage and current, and then average (or in the case of a kWhr. meter, integrate) the result. Charlie Sullivan charless@cory.berkeley.edu
rsd@sei.cmu.edu (Richard S D'Ippolito) (01/03/91)
In article <10003@pasteur.Berkeley.EDU> Charles Sullivan writes: >>Your KWHr meter has always been RMS-responding. >>[...] >>Rich >RMS is not and should not be used in 'true' power meters. They read just >plain average power ( I will use '<P>' for this). The only reason people >use RMS to begin with is to be able to calculate <P> from V^2/R by using >V=RMS(V(t)). Indeed, Mr. Sullivan is correct and my short statement about the KWHr meter is incorrect. The mechanical and hall-effect electronic meters do multiply the instantaneous values and integrate. Sorry for its inclusion. Rich
dam@cs.glasgow.ac.uk (Mr David Morning) (01/04/91)
I've never really understood why the term "root mean square"(RMS) is applied to power measurements. Power = (V)rms*(I)rms would imply to me that 2 roots are being mutiplied together leaving just a mean squared value. Similarly Power= (I)rms^2 * R cancels the the "r" of the "rms" I used to work for a firm which manufactured AC digital panel meters. The only time I've heard the term "True RMS" used was in relation to a digital multimeter that incorporated a true RMS chip. This chip prformed all sorts of convoluted analogue maths to arrive at the answer. Might have used Simpsons rule or something similar but all the other AC meters were termed "Average Sensing, RMS scaled". Dave "