With Wings As Eagles: Craig P. Steffen's Blog

average power (RMS power is right out)

2009 December 26 07:53

I used to refer to the "RMS power" of something.  I thought it was a reasonable term and a specific quantity.  Turns out that's not the case.  Here's an article that sort of explains reason that's not true. 

Here's my attempt to try to explain the distinction:

The instantaneous electrical power consumed by a device is the product of the voltage across the input (energy per unit charge) times the current flowing (current charges per second).  Multiply those together and you get energy per time, or power (in watts).

If the voltage and current are constant over time, the the power consumed is just the product of the voltage times the current.  

If the voltage and current change with time, then the power may (and usually will) change over time.  You calculate power consumed at a given instant by multiply the instantaneous voltage times the instantaneous current at that same instant.  What's often more useful is to calculate the average power used over time.  Mathematically, this involves taking a number of correlated voltage and current samples over a time period, summing the products, and then dividing by the number of samples.  The result is the average power consumed over that period.  

Now, there is an important simplification that can be applied ONLY WHEN  the input voltages and current are sinusoidal and IN PHASE.  A weighted voltage and a weighted current can be computed separately, and their product equals the average power drawn. This is simpler to calculate because you're only mutiplying V*I once, not for each moment in time.   These weighted values are called "RMS" values which stands for "root-mean-squared".  RMS is a simplification to voltage and current separately because they're sinusoidal.  There is no "RMS power" calculated; no such thing exists.  It's just "average power".