[Prev][Next][Index][Thread]

Re: resistance in an LRC circuit used to calculate time constant



Tesla List wrote:
> 
> Original Poster: "Alfred C. Erpel" <alfred-at-erpel-dot-com>

> So far, what I have read regarding time constants is (one time constant) =
> L/R for inductance and (one time constant)= RC for capacitance. Could you
> explain why in the context of the series resonant circuit you described
> above, that 1t = 2*L/R and that it is regardless of C? Are you also saying
> that RC is irrelevant in this circuit under any circumstances? If it is not
> too involved, please explain.  Thank you.

I will try, but this requires some calculus with Laplace transforms and
some circuit theory. (I could also use differential equations, and much
greater complication.)
 
Suppose the following model:

             I
     +----L-->-+
  +  |         |  
 Vc  C         R  
  -  |         |  
     +---------+

Vc(0) is the initial voltage at the capacitor. It can be modeled as a
step
voltage Vcs in series with an uncharged capacitor. 
In Laplace transforms:

Vcs(s) = Vc(0)/s.

The current in the series circuit is this voltage divided by the sum
of the three impedances in the circuit. In Laplace transforms:

Z(s) = 1/(s*C)+s*L+R

I(s) = Vcs(s)/Z(s) = Vc(0)/s/(1/(s*C)+s*L+R) = Vc(0)/(1/C+s^2*L+s*R) =
=(Vc(0)/L)/(s^2+s*R/L+1/(L*C))

This is in the form:

I(s) = k/((s+b)^2+a^2)

that, inverting the Laplace transform, corresponds to the waveform in 
the time domain:

I(t) = (k/a)*exp(-b*t)*sin(a*t)

where:

k=Vc(0)/L
a=sqrt(1/(L*C)-R^2/(4*L^2))
b=R/(2*L)

The current is a sinusoid with a frequency that is -approximately-
(I forgot this) 1/(2*pi*sqrt(L*C)), decaying with a time constant 2*L/R.
This is valid as long as "a" is positive.

C doesn't affect the term multiplying s in I(s), and so disappears
from the time constant of the exponential decay.

The complete theory about this is normally found in books about circuit 
theory.

Antonio Carlos M. de Queiroz