Workaround to measure RMS on a non-rms meter

[percy]

My DMM does not measure rms ac. I was thinking about this workaround "probe" that allow me to do that until I get one with will measure true rms. So I just wanted to run this by you guys to see if you thought if the idea will work or not.

I am thinking of just wiring up a bridge rectifier with a 1000uf filter cap, apply the ac signal that I want to measure the rms for at its input, and measure the filtered DC with the dmm.
The rms voltage of the input ac signal would then be Vrms = Vdc/1.1414 -Vd (i.e. Vdc divided by 1.414 minus the diode drop). Do you think this method is ok ?

[zinsula]

If you want to measure a voltage, you may want a high input impedance of your rig. 1000µF is not a high impedance until it's charged. You may even damage your DUT.
But then, it depends on what you want to measure.

A better solution would be a RMS to DC converter.

http://www.analog.com/en/prod/0,,773...D737%2C00.html

Success, Tino

[sreten]

Originally posted by percy
My DMM does not measure rms ac.

Are you saying it measures AC but not rms AC ?

/sreten.

[jackinnj]

if you are measuring a sine-wave, then the RMS voltage is 1/2 the square root of 2 times the peak voltage.

if you are measuring noise, music etc. you can, as ZIN recommends, use an Analog Devices true rms chip like the AD536 or AD737 -- (the former is used in the Hewlett Packard HP3478 and HP3468 DVM's and in the Boonton 1120 distortion analyzer) -- but you need a high quality (polycarbonate or film) averaging capacitor. The AD737 is a low voltage device and needs some protection, it costs ~$6 from DK, the AD536 is about $12.

the most acurate RMS measurement techniques recently developed use high speed ADC's -- they are finding applications in power measurement for utility companies now that energy is so expensive.

Other than that the HP 3400, 3403 true RMS meters are cheap as dirt on EBay -- they use a thermopile to measure temperature and convert to RMS voltage -- every audio workbench should have a decent RMS voltmeter. Used to be that Linear Tech made a calorimetric chip, the LT1188 I think -- but it has been discontinued -- it required quite a bit of surrounding circuitry.

oh -- and shameless plug -- I have AD737's and AD536's on my webstore, along with some very high quality capacitors suitable for use as C(avg).

[percy]

yes it does measures AC but its not rms and I am not sure what it is (peak, average,..). The manual says it measures "average-rms"(?!)

For now I just want to measure rms voltage of audio signals (in mostly power amplifiers) in the 1 to 25V range.

[sreten]

Hi,

as it measures "average rms" AC it will do a better job
than any simple work round that you could come up with.

A true RMS AC meter can measure any repeating waveform,
(e.g. 1 in 10 duty cycle square wave) not just a sine wave.

To confirm what it measures measure the mains voltage.

What do you mean by audio signals ? If music you will
need true RMS, if sine waves your meter should be fine.


/sreten.

[percy]

well it measures 118 to 120VAC (U.S) in the household line, but what does that tell ? What's average rms ?

No not music, just sine waves, maybe warble. But what about MLS though ?

Another issue is frequency of the signal. I don't think its accurate above ~400hz. So if I want to measure 1Khz it useless anyway.

I will find my notes and post the exact voltages (ac and dc) I measured with a basic power supply. Thats when I thought something's not right in the ac voltage measurement. That will give you an idea.

<edit:> I know I need a new meter. This one's even so slow. I was just wondering if there was something quick I could do.

[Nigel Goodwin]

A 'normal' meter on AC simply uses a rectifier to convert from AC to DC, the scale is then calibrated to read the correct RMS value for a sinewave - for any other waveform it will read incorrectly. As you are also aware, they only read correctly for a very limited frequency range, basically being designed to read mains voltages.

A true RMS meter mathematically converts different waveforms to give the correct readings, specific IC's are available to do this. However, I suspect that they may well be intended for limited frequency use.

What are you wanting to read that isn't mains anyway?, generally an AC milli-voltmeter would be used for audio uses.

[sreten]

hi,

you do need a new meter (or an oscilloscope ).

"average RMS" I believe is just a simple scaling of rectified peak values.
Which will obviously be inaccurate for complex signals.

You are also right regarding ineffectivenees at high frequencies

/sreten.

[EC8010]

What makes you think you need to measure RMS? Unless a meter specifies true RMS, it measures average but is calibrated RMS of sine wave, that's what your "average-RMS" means. The only time I've needed a true RMS meter was when I was unavoidably feeding heaters with a distorted sine wave and wanted to be sure that they received the correct heating power. It's traditional to measure RF noise in RMS, but you wouldn't use a DVM for that. Audio noise can be measured RMS if you need a flattering figure, or quasi-peak if you want something that more accurately reflects the nuisance.