pn diode step response using delta generation



I’m trying to get a simple transient step response simulation using delta generation sources. I place several delta generation sources in the pn junction diode, configure a transient simulation, and apply a global shutter to create a current step input. I set voltage tables on the p-type contact to define the bias voltage and the simulation time duration. I set the minimum time step to 0.1 fs and the maximum time step to 100 fs (should be overkill). I define the step input to start at 5 ps with a 1 ps slew, and the simulation runs for 20 ps.

The simulation successfully runs, but the transient current through the p-type contact follows the input current source seemingly exactly – i.e. it is indistinguishable from the slew of the step input and increases to its maximum current value in 1 ps. This is extremely unlikely to be accurate.

I have attached the script that creates and runs the simulation. Any ideas as to what I’m doing incorrectly?

transient_diode_step_response.lsf (4.4 KB)


Hi. I believe the problem (wrong result) was coming because the CHARGE solver was using the “gummel” solver. While running transient simulation you should always choose the “newton” solver. If you simple set the solver type to “newton” in the CHARGE solver properties, you will see the proper time evolution of the contact current.

setnamed("CHARGE","solver type","newton");


Great, this seems to work! Thank you for the response.