What does "TRUE RMS" mean?

[mike]

I have a TRUE RMS DVM.
Hooked it in parallel with a digital scope to the
function generator at 60 Hz.
At zero offset, the DVM reading tracks the scope RMS
calculation for sine/square/triangle waveforms.

The scope knows how to measure the RMS value of a
sinewave with voltage offset.
4V DC is 4V RMS on the scope, but zero on the DMM.

A look at the DVM specs shows that the AC RMS measurement
has a low frequency of 45 Hz.
That explains the observation.
BUT
That does not sound like a TRUE RMS measurement????

 

[Spehro Pefhany]

I have a TRUE RMS DVM.
Hooked it in parallel with a digital scope to the
function generator at 60 Hz.
At zero offset, the DVM reading tracks the scope RMS
calculation for sine/square/triangle waveforms.

The scope knows how to measure the RMS value of a
sinewave with voltage offset.
4V DC is 4V RMS on the scope, but zero on the DMM.

A look at the DVM specs shows that the AC RMS measurement
has a low frequency of 45 Hz.
That explains the observation.
BUT
That does not sound like a TRUE RMS measurement????

Well, it's better than the usual diddled (*1.11) average value that
you see.


Best regards,
Spehro Pefhany

 

[Phil Allison]

"mike"

I have a TRUE RMS DVM.
Hooked it in parallel with a digital scope to the
function generator at 60 Hz.
At zero offset, the DVM reading tracks the scope RMS
calculation for sine/square/triangle waveforms.

The scope knows how to measure the RMS value of a
sinewave with voltage offset.
4V DC is 4V RMS on the scope, but zero on the DMM.

A look at the DVM specs shows that the AC RMS measurement
has a low frequency of 45 Hz.
That explains the observation.
BUT
That does not sound like a TRUE RMS measurement????

** True RMS means that within the specified frequency limits and max crest
factor the meter responds to the rms value of the wave.

This is quite unlike most analogue meters and DMMs which respond to the
*average rectified value * of the wave and display a reading that is
adjusted to give correct rms values only for sine waves.

Budget DMMs that display "true rms" have very limited frequency range ( ie
45Hz to 1000Hz ) and modest max crest factors like 3. They are useful the
waves found in mains frequency PSUs.

Better DMMs may work accurately from DC to 100kHz or more and crest factors
of up to 7.

The best use thermal methods and go out to GHz.


.... Phil

 

[P E Schoen]

"mike" wrote in message

I have a TRUE RMS DVM.
Hooked it in parallel with a digital scope to the
function generator at 60 Hz.
At zero offset, the DVM reading tracks the scope
RMS calculation for sine/square/triangle waveforms.

The scope knows how to measure the RMS value of a
sinewave with voltage offset.
4V DC is 4V RMS on the scope, but zero on the DMM.

A look at the DVM specs shows that the AC RMS
measurement has a low frequency of 45 Hz.
That explains the observation.
BUT
That does not sound like a TRUE RMS measurement????

Many true-RMS meters use AC coupling to remove the DC component, but good
ones also have an AC+DC option, which is simply DC coupled. The AC coupling
is very useful for things such as ripple on a DC signal. But the AC+DC
measurement is necessary for things like variable duty-cycle rectangular
waveforms and PWM signals. Generally you can determine the actual true-RMS
value of a signal with DC offset by using Vrms = sqrt( Vac^2 + Vdc^2).

In my work, I need to measure the true-RMS value of a short pulse of power
line frequency AC. If you do a computation of the value based on samples,
the value will vary widely for short pulses but then converge eventually to
the steady-state true-RMS value. I have found that it is better to use a
sampling rate that is an integer multiple of the line frequency, so 1200 Hz
works well for both 50 Hz and 60 Hz, with 24 or 20 samples per cycle. And
when reading the RMS value, a sample of 100 or 200 mSec shows the least
amount of "jitter" since it comprises exactly 5 or 6 complete cycles.

This is also SOP for many DC digital meters, so that AC line noise is
essentially canceled out, resulting in relatively flicker-free readings.

Some discussion of this can be found in an article I wrote on Circuit
Breaker Testing Technology: http://www.pstech-inc.com/cbtest.htm

Paul

 

[Phil Allison]

"Dorothy with the Red Shoes on"

What the meter maker means by "True RMS" is that the meter is better
than the old, cheap "RMS" meters which simply averaged the value from the
peaks. The new meters actually analyze the waveform from moment to
moment.

** DMMs ( hand held or bench types) with "true rms" ranges date back to the
early 1980s when Analog Devices introduced their " true rms to DC" converter
ICs.

A DMM is essentially an analogue meter with a numerical display based on a
pulse counter and dual slope integrator. This arrangement can only measure
DC voltages. In order to read AC ones, there is an AC to DC converter -
either a precision rectifier or a true rms to DC converter inside the DMM.
This is engaged when an AC range is used.

It is the performance of the latter circuits that determines the AC
performance of the meter - plus the design and bandwidth of the input
attenuator.

The effective sampling rate of most DMMs is only 3 or 4 times a second - ie
the same as the display update rate.



.... Phil

 

[Phil Allison]

"Phil Hobbs"

One of the earlier digital true-RMS meters is the HP 3403A, which came out
in about 1972. It has AC, DC, and AC+DC modes.

** A thermal voltmeter - not a DMM.

I mentioned the type in my previous post in this thread.

But the HP3403a is only good to 10MHz.

http://www.hpl.hp.com/hpjournal/pdfs/IssuePDFs/1972-03.pdf

** Way too big.


.... Phil

 

[[email protected]]

One of the earlier digital true-RMS meters is the HP 3403A, which came
out in about 1972. It has AC, DC, and AC+DC modes.

http://www.hpl.hp.com/hpjournal/pdfs/IssuePDFs/1972-03.pdf

The early TrueRMS and wide bandwidth meters even today are essentially
(RF)dummy loads with a thermometer.

It is interesting to note that English people talk about RMS (Root
Mean Square) values, while in many European languages something
equivalent to effective value (e.g. Effektivwert) is used, i.e. the DC
level that would produce the same heating effect in a resistive load
as the AC/DC signal being measured.

 

[Phil Allison]

"Phil Hobbs = PITA troll "

It's digital, anyway.

** You are being a smartarse prick - as bloody usual.

The HP3403A has NOTHING to do with my post about DMMs with AD rms chips.

Takes the same amount of shelf space

** ROTFL !!!

The pdf FILE is too fucking BIG - idiot.



.... Phil

 

[Phil Allison]

"Nutcase Turd

Arent the old fashioned 2 coil meters rms?


** That sounds like a power meter.

Got SFA to do with my post about DMMs.

Can you read at all ??



..... Phil

 

[[email protected]]

I don't know about that. True RMS meters theoretically don't make a lot
of assumptions about their input signals, whereas there's a very long
history of measurements based on peak- or average-reading AC meters with
their scale dorked to read the correct RMS value, assuming that the
input waveform is sinusoidal.

For noise waveforms, average-reading meters read 1 dB low--a little
matter of 10% error.

"Electrical equivalent" would leave me wondering what the underlying
assumptions are. "True RMS" is a lot clearer.

I think effective value is more intuitive.

Look at the HP link at page 17, what this circuit actually does, could
be described with the following analogy.

Two identical 50 ohm dummy loads with identical thermometers are used.
The (modulated) RF signal to be measured, is fed into the other dummy
load and the temperature is observed.

A DC supply is connected to the other dummy load through a
potentiometer and the potentiometer is adjusted, until the other
thermometer reads the same as the first thermometer. The potentiometer
output voltage is measured using any ordinary DC volt meter. The DC
voltage reading is the effective value of the modulated RF waveform.

 

[Phil Allison]

"Nutcase Turd"


I've seen the phrase 'EE' used on 1930s equipment here, meaning
electrical equivalent. Its a lot more intuitive than rms.


** Its totally ambiguous while " rms" is very precise.

The term comes from statistical analysis - the "root of the mean square "
of a series of numbers.

The physics connection is pure serendipity ....



.... Phil

 

[Cydrome Leader]

Sure, it's called electrical substitution, and it's been used in optical
power meters since before the invention of rocks. I understand the
concept perfectly well, I just disagree with you about the clarity of
the nomenclature. "True RMS" gives me a lot more confidence that the
meter is doing what I think it is.

That is if you trust the meter or what it claims it can do.

A friend got their first meter- one of those dollar store looking ones.
I flat out told them, don't even try to use this to test an outlet.

I think it had a 600 volt AC range too. I think the leads were the
fuse. I would not be surprised if it claimed to be a true RMS meter.

 

[Ralph Barone]

Sure, it's called electrical substitution, and it's been used in optical
power meters since before the invention of rocks. I understand the
concept perfectly well, I just disagree with you about the clarity of the
nomenclature. "True RMS" gives me a lot more confidence that the meter
is doing what I think it is.

Cheers

Phil Hobbs

Except that whenever I see the phrase "True xxx", I automatically assume
that someone is lying:-(

 

[Spehro Pefhany]

Most meters still do it that way. Seems to me that it would be better,
these days, to use a fast sampling ADC, digitize whatever's there, and
do the math.

John

I'm sure AD could make a chip with a DSP on it that would do just
that, just as they effectively incorporate DSPs into some of their
high-end ADCs.

Not sure there's much demand for really high accuracy RMS measurements
though... for the ones I've done a few tenths of a percent is gross
overkill.

 

[Spehro Pefhany]

The file is 4Mbytes - hardly "big" even if you're on dial up.

Maybe Australia needs to upgrade their modem.

 

[Cydrome Leader]

You mean, Won Hung Lo meters aren't as good as Flukes? Who knew?

I have a really old fluke and wavetek meter that I trust. I have to do
the sanity check when I use anybody else's meters though. A no-name
meter with old batteries is always sketchy.

It would be nice if meters had an indicator for "you blew your current
ranges".

 

[Phil Allison]

"Phil Hobbs = PITA troll"

You appeared to be claiming that true-RMS digital voltmeters didn't exist
before the AD chips.

** That is demented.

Try learning to read - you pathetic wanker.

Here is the first line:

" DMMs ( hand held or bench types) with "true rms" ranges date back to the
early 1980s when Analog Devices introduced their " true rms to DC" converter
ICs. "

And follow the CONTEXT - you bullshitting autistic pig.


.... Phil

 

[Phil Allison]

"Nutcase Turd

Arent the old fashioned 2 coil meters rms?

** That sounds like a power meter.

correct. AIUI they do a 4 quadrant multiplication followed by
averaging.


** Got SFA to do with my post about DMMs.

Can you read at all ??

You bullshitting pile of dung.

 

[Phil Allison]

"John Larkin"
"Phil Allison"

Most meters still do it that way. Seems to me that it would be better,
these days, to use a fast sampling ADC, digitize whatever's there, and
do the math.

** Try doing the costing and making it work on a 9v battery at under 1mA
drain.

DMMs with " true rms " sell for under $50.


.... Phil

 

[Phil Allison]

"Nutcase Turd"



** **** off - you pathetic, bullshitting dickhead.