On Sat, 18 Jun 2005 06:42:29 -0700, cs_posting wrote:
ehsjr wrote:
wrote:

The L and the C don't care about your DC offset, so you must still
think of the signal as AC in order to understand their behavior. They
don't care that the overall signal doesn't reverse polarity, they only
care that derivative of voltage with respect to time is non-zero.


Er - there are cases where the L will be saturated by the DC component.

What you are suggesting is a good issue to keep in mind for the real world
(and one I had overlooked).

However, what you have actually said is not true.

An inductance - a specific element we both referred to as L - will not
saturate. Rather it will behave in accordance with the simple
mathematical model of inductance.

The real-world magnetic device chosen to play the role of an inductor can
saturate, and it's something we might need to think about. However the
propensity towards saturation would need to be specified by additional
parameters beyond a simple constant value of L. While we're at it, we
should put in parasitic resistance, temperature dependence, possible
effects of external fields, and probably some other things that I'm not
thinking about.

If asked to solve a problem with an inductance, you treat it as such. If
asked to solve a problem with an inductor, you have to consider the
broader properties of that device, of which inductance is only one, and
not necesssarily a constant one.

I remember learning this very thing at a bench in USAF electronics tech
school. We did the numbers on paper, and then hooked up a real circuit
and all of the phases were off from what we were expecting from the
numbers because we had neglected the inductor's own resistance. Once
we put that back into our equations, of course, it all came out right. :-)

Cheers!
Rich