True-rms meters must be more complicated than many would imagine, especially
if they have to deal with extreme types of wave forms.
Consider a 1 volt DC signal. This has a true-rms value of 1 volt. Consider
a sinusoidal signal with a peak voltage of 1.414 volts. This also has a
true-rms value of 1 volt. If all signals were this simple things would be
easier.
Now consider a square wave switching between zero volts and 2 volts with a
duty cycle of 50% (this means it spends 50% of its time at zero volts and
50% of its time at 2 volts). This also has a true-rms value of 1 volt.
Likewise a 10% duty-cycle square wave with peak value of 10 volts has a
1-volt true-rms value.
Manufacturers of true-rms meters are attempting to make the meters read 1
volt in all these cases. To do this requires active circuitry using
transistors, etc. Here are some problems:
(1) As the duty cycle gets small, the peak value must get high. If the duty
cycle of a square wave is 1%, the peak value must be 100 volts to yield a
true-rms value of 1 volt. So we're going to set the meter range to 1-volt
AC, but feed it a signal that peaks at 100 volts. Can the circuitry handle
that? Maybe not. To protect our transistors we would really like to put
that 100-volt square-wave signal through a voltage divider to reduce the
peak value presented to the circuitry, but then the duty cycle is so small
that the rms-value coming out of the voltage divider is too small for the
range we would like to be using.
(2) Active circuits also have lower limits on the signals they can process.
Internal noise in the electronics means measuring very small rms values is
tough. One source (URL given below) indicates the meter can give an
internal noise reading of several digits if you are trying to measure an RMS
level that is 1% of the range you have selected. This might be some of what
people are seeing when they torch resistors.
Some true-rms meters are "AC-coupled," e.g. they have something like
blocking capacitors to block any DC component of the inputted signal. In my
mind this means they are not true-rms. If we have a 50% duty-cycle
1-volt-peak square wave riding on top of a 1-volt DC level, I would say the
true-rms value is 1.5 volts. That's what a DC-coupled meter ought to read.
But an AC-coupled meter will read 0.5 volts.
The question is "what do you want?" Do you want the rms value of only the
AC component, or do you want the rms value of the whole signal. Both could
be valid, and some meters will do one while other meters will do the other.
Michael D. Edmiston, Ph.D. Phone/voice-mail: 419-358-3270
Professor of Chemistry & Physics FAX: 419-358-3323
Chairman, Science Department E-Mail edmiston@bluffton.edu
Bluffton College
280 West College Avenue
Bluffton, OH 45817