Definition of current ripple

Question

I understand the definition of the ripple factor (for current) as \$ripplefactor = \frac{\sqrt{I_{rms}^2-I_{avg}^2}}{I_{avg}} \$

Is this the same as "current ripple" (in percentage) where the context is output current of a thyristor converter?

Answer 1

Taking into consideration a DC distribution network and a pair of applicable standards (EN 50155 for DC network onboard rolling stock and MIL STD 704 for distribution onboard aircrafts) the definitions of ripple (for voltage, equivalent for current) are:

1. peak-to-peak variation with respect to steady state value (your Iavg) : Ipp/Iavg
2. peak-to-steady value variation, but has the problem of asymmetrical variation above and below this steady value, so it's difficult to apply;
3. rms variation with respect to steady state value : Irms/Iavg

I have focused in the past on voltage ripple as a definition of delivered DC voltage quality. However, definitions and considerations may be applied to current as well.

Definition 1 is a time-domain definition where instantaneous values are taken so that the high and low peak might occur one immediately after the other, or slowly very well separated. It is apparent that these are two different phenomena (a spike ad a fluctuation) but the definition of ripple would not be able to distinguish.

Definition 3 may be implemented in time or frequency domain. It is easy to see that you can calculate the rms of selected components, giving a concept of bandwidth (e.g. up to the 40th harmonic).

Harmonic of what?! ... we are in DC! The problem in fact is that ripple definitions have been offered with an AC mains in mind, and then DC appear after a rectifier, so that you have some low-order characteristic harmonics (e.g. 2nd, 6th, ...).

Bandwidth or dynamics may be included also in definition 1 in time domain: only you take the peak-to-peak of values that are separated by at least tmin seconds and no more than tmax seconds.

So, Irms/Iavg does not correspond to your definition, nor the normalized peak-to-peak value.