Numerical solution of a differential equation with time dependent terms

5 replies · 2K views · last reply
Carlos Criollo
Messages
9
Reaction score
0
I am would like to solve this differential equation:

Typeset differential equation dV_c(t)/dt = (1/(Z_a C(t))) V_bias - (1/(Z_a C(t))) V_c(t) - (β(t)/(Z_a C(t))) V_c(t) - (G_s(t)/C(t)) V_c(t) - (1/C(t)) (dC(t)/dt) V_c(t).


Where

Equation 4: G_s(t) = (W/L) e mu_e I_l exp(-2)(1 - R)(1 - exp(alpha T_LT-GaAs)) times sqrt(2 pi)/(4 h v_opt) tau_l exp(-tau_l^2/(8 tau_c^2) - t/tau_c) times (erf(sqrt(2) t/tau_l - sqrt(2) tau_l/(4 tau_c)) + 1).

Equation: C(t) = (τ_r / Z_a) (1 + (e μ_e Z_a S n(t)) / L), where C(t) is a function of time t.

http://ieeexplore.ieee.org.ezproxy.uniandes.edu.co:8080/ielx5/8/6493417/6409989/html/img/6409989-eqdisp-3-small.png
Equation: beta(t) = e mu_e n(t) tau_r / (zeta epsilon), where beta(t) is a function of time, e and zeta are constants, mu_e is a mobility, n(t) is a time-dependent quantity, tau_r is a time constant, and epsilon is a permittivity.


Could you give me some practical ideas about the required software and methodology? Thank you very much
 
If all the parameters are known, did you try to just calculate it time steps by time step? Did it work?
There are more precise integration methods, but that is something you can have a look at afterwards.
 
Yes, all the parameters are known, but how I can calculate the time step by time step?
 
Carlos Criollo said:
Yes, all the parameters are known, but how I can calculate the time step by time step?
Time is the independent variable. You start your calculations at t = 0, increment t, rinse and repeat.
 
And how I could implement it in software?, I need a numerical solution of this equations.