I was working on a little project and I was wondering if the envelope detector circuit could be resolved analytically. Let me explain.

Envelope Detection System

The circuit is the one shown above, I know the values of $C$, $R$ and $V(t)$ which is the input signal. My $V(t)$ is a smooth function and I know all of its derivatives. The thing I would like to find is the function of time $V_d(t)$ which is the voltage drop across the diode assuming the Shockley equation to be true at every instant in time.

$$ I(t) = I_s \Big(\exp \Big(\frac{V_d(t)}{V_t}\Big) - 1 \Big)$$

Another thing that I assume is that Kirchhoff Voltage law is true for every $t$. I've found a pretty nasty differential, first order, equation, which is:

$$ \exp\Big(\frac{V_d(t)}{V_t}\Big) = \lg \Big[\frac{1}{I_s} \Big( \frac{V - V_t}{R} + C\frac{d}{dt} (V - V_d) \Big) + 1\Big] $$

Can this circuit be solved? Does the result has any limitation other than the ones mentioned above?

EDIT: I've realized that the question is meaning less without knowing the $V_t$, in this case it is: $$V_t(t) = A(1 + m \cos (2 \pi f_1 t + \phi_1))cos(2\pi f_2 t+ \phi_2)$$


The differential equation can be written in the form :

$\frac{d{{V}_{s}}}{dt}=-\frac{{{V}_{s}}(t)}{RC}+\frac{Is}{C}\left( {{e}^{(V(t)-Vs(t))/{{V}_{d}}}}-1 \right)$

I can be wrong but I do not think she admits an exact analytical solution.

But this differential equation can be solved numerically very easily (with Python for example)

| cite | improve this answer | |
  • $\begingroup$ I thought so, I can easily solve the homogeneous equation but I can't find any solution for the non-homogeneous one, due to the form of $V(t)$. $\endgroup$ – superporchetta Feb 22 '19 at 11:07
  • $\begingroup$ The equation is non linear. So it is not possible to separate the homogeneous solution and the complete one. $\endgroup$ – Vincent Fraticelli Feb 22 '19 at 11:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.