What should the voltage be across phases on a 3-phase supply (VFD)?

[Matrixx]

The output from a VFD will not be measured remotely accurately by an analogue meter, they are designed to drive motors, which are driven by frequency (on an AC motor). I have found a good Multimeter like a fluke could read the voltage , but analogue meters could not, the output is all over the place, the noise would normally be absorbed by the windings of the motor, and most have vector feedback so it knows the speed of the motor.

To get analouge meters to work I had to use a three phase filter to absorb all the harmonics/ noise etc, I was using the VFD to simulate a generator.

 

[Maker]

It's auto-ranging and auto DC/AC.

Is it any good though? "Auto AC/DC doesn't inspire confidence . A decent meter will give the right voltage on any AC waveform (Fluke call it True-RMS), cheap ones just take 30 percent off the peak value and call it good enough but for anything but a true sine wave, it's miles off good enough.

Analogue meters have a "time constant" associated with them due to them having some small inductance and capacitance, anything that happens very quickly (A PWM pulse for example) is likely to be missed, or greatly smoothed in a way that may or may not be useful. I wouldn't touch an analogue meter, I can't believe anyone still uses them.

 

[Matrixx]

Is it any good though? A decent meter will give the right voltage on any AC waveform (True-RMS, I think that's a Fluke trademark), cheap ones just take 30 percent off the peak value and call it good enough but for anything but a true sine wave, it's miles off good enough.

Its not a symmetrical wave so RMS won't work, I have seen the wave on an oscilloscope and its just a mess.

 

[daleyd]

In 20 odd years of working with drives I have never had a problem measuring output volts when using a decent meter.

Some screen caps from a fluke meter, showing incoming volts and output volts at 50hz and 25 hz.

Out of interest I got a cheap meter I have in the garage and it wouldn't read at all on the output side - measured the input volts identical to the fluke one .....

 

[Maker]

Its not a symmetrical wave so RMS won't work, I have seen the wave on an oscilloscope and its just a mess.

That shouldn't matter, a proper RMS measurement should be unaffected by harmonics and strange waveforms. I'm not sure how it's done in a handheld meter (It can't just average loads of readings at a high sample rate can it?). High end transfer standards and calibrators use a thermal sensor where the waveform under test heats a resistor and a DC voltage is adjusted to heat an identical resistor at the same rate, I'm pretty sure that's far too posh to be in every multimeter though.

 

[awemawson]

If you REALLY want the true RMS of an odd waveform use a "hot wire ammeter", which is why the output stages of military transmitters were measured thus in the 40's and 50's