Averaging RMS Voltage Over Time

[RMSquest]

If I have multiple RMS voltage values over a day's time and I want to find the average - should I use RMS or arithmetic mean?

I think I would use arithmetic mean because RMS is just to account for the +/- direction of voltage, but I'm looking for confirmation.

Thanks!

 

[vk6kro]

If I have multiple RMS voltage values over a day's time and I want to find the average - should I use RMS or arithmetic mean?

I think I would use arithmetic mean because RMS is just to account for the +/- direction of voltage, but I'm looking for confirmation.

Thanks!

You should collect the RMS values and take an average of them over a day or whatever time you like.

The RMS process not only allows for the alternating polarity, it also allows for the fact that the heating effect of a waveform depends on the square of the instantaneous voltage.
So, the top of a sinewave has a much greater effect on the heating than the lower parts. More than you would expect from just the voltage. This is the main advantage of using RMS, especially for waveforms that are not sinewaves.

RMS is the DC voltage that gives the same heating effect as the waveform you are measuring.

If you are sure that the waveform is a sinewave, it is possible to measure average voltages allowing for the polarity and then calibrate this to RMS. Cheaper multimeters have been doing this for years.

 

[sophiecentaur]

You should collect the RMS values and take an average of them over a day or whatever time you like.

The RMS process not only allows for the alternating polarity, it also allows for the fact that the heating effect of a waveform depends on the square of the instantaneous voltage.
So, the top of a sinewave has a much greater effect on the heating than the lower parts. More than you would expect from just the voltage. This is the main advantage of using RMS, especially for waveforms that are not sinewaves.

RMS is the DC voltage that gives the same heating effect as the waveform you are measuring.

If you are sure that the waveform is a sinewave, it is possible to measure average voltages allowing for the polarity and then calibrate this to RMS. Cheaper multimeters have been doing this for years.

If you want to get a meaningful result from 'averaging' your RMS values then you really need to know the length of time over which the 'M' in 'RMS' was calculated or assumed. The above answer seems to assume that we are dealing just with normal AC mains, which is just one example of where RMS is used.

As long as you are taking RMS readings at regular intervals then taking their Mean is OK but, if the readings are not taken at regular intervals, a simple Mean calculation may not give the best (most representative) result because the readings will each have different significance. Imagine, for instance, that you took a burst of readings when the volts happened to be a bit high and then took regular, wide spaced, readings over a longer period. The extra lot of high readings in your simple Mean calculation would bias the result in the 'high' direction. So, in that case, it would be more reliable to weight the readings by a factor corresponding to the time between readings. This may or may not be relevant to the particular case in the OP but it is worth while pointing out.