Precise Simulation of Pulse Signal of a Diode Detector

Simulation / Configuration Questions
2

Hi @Andrew

Unfortunately I have no access to the repository:

image

It’s hard to pinpoint potential issues from the information I have available just now. Could you elaborate a bit (or post the configurations you are running? A few things to consider:

  • We’re dealing with MC simulations, so we are subject to fluctuations. When we look at pulse shapes, we always generate several pulses (a few tens or hundreds) from the same starting position, and then average those to get a nice pulse
  • Make sure to use a small time stepping, depending on the detector something like 10ps might be worth looking at.
  • Make sure the weighting potential you are using is matching the electric field. How does the weighting potential and electric field maps look like that the respective readers produce for you?
  • Does the integrated charge match your expectation? How much do you deposit initially? Are you sure the time axis doesn’t cut off the actual pulse?

An example from a recent simulation from our group comparing a TCAD simulation with an (averaged) APSQ simulation using the electrostatic TCAD fields:

image
image1918×971 12.9 KB

What we use for this is (abbreviated, relevant parameters remaining):

[DepositionPointCharge]
source_type = "mip"
model = "fixed"
position = 30um 30um 
number_of_steps = 30
number_of_charges =60/um

[ElectricFieldReader]
# ...

[WeightingPotentialReader]
# ...

[TransientPropagation]
temperature = 293K
charge_per_step = 1
timestep = 10ps
timestep_max = 10ps
induction_matrix = 1 1
mobility_model="masetti_canali"
integration_time = 40ns

[PulseTransfer]
max_depth_distance = 0.01um
collect_from_implant = true

[DefaultDigitizer]
electronics_noise = 0e
threshold = 100e

I hope this gives you some hints, please feel free to come back with your configs for discussion!

Cheers,
Simon

2 Likes