Mismatch between values from an imported electric field (TCAD) and Allpix2 mesh_plotter?

1

Hi! Very new to this, so please bare with me :slight_smile:

I have a basic toy TCAD simulation of a biased (~3V) sensor (100um thick, 25um epi with pixels of 28x28 um^2) with some reasonable guesses of the doping profiles based on information available from online publications.

The conversion of the field goes without problems using the mesh_converter (although it takes quite a bit of time to compile). The options I used for this are:

region = "region1"
observable = 'ElectricField'
observable_units = 'V/cm'
divisions= 200 2000 200
xyz = z,y,-x

The results of the mesh_plotter I have to admit I don’t really understand, as I cannot replicate the correct values of the electric field. The shape seems correct however. Here are the results from the interpolation on yz without additional parameters (CERNBox)

I shoot 100000 e- of 5.4GeV in a telescope-like setup with [DepositionGeant4]. The import of the Efield happens as here:

[ElectricFieldReader]
type = "alpide"
model = "mesh"
file_name = "ElectricField.apf"

Furthermore, I use [GenericPropagation],[SimpleTransfer],[DefaultDigitizer] and [DetectorHistogrammer] with good guesses to check what happens. I cannot replicate the correct cluster size histogram from data, in the sense that I have very few clusters >2px.

I would like to ask if my understanding of the import of the Efield is correct.
Did I correctly understand the change of coordinate system? Or should it instead have a -x in AP2 coordinate system? Should the unit cell (defined as one pixel in TCAD, with the diode in the center, as can be seen in the provided picture) be offset by 0.5,0.5?
Does the Efield look reasonable? Why are the absolute values not interpreted correctly? I would have expected some values in the order of several 10^3 V/cm…

2

Hi @mblidaru

These are plotted in framework-internal units. Plugging together the relevant ones from the manual this adds up to being MV/mm, which for me adds up:

4e4 V/cm / 1e6 / 10 = 0.004 MV/mm

which is what you got in these plots.

I’d suggest adding

output_plots = true

so you see the final result loaded for the simulation. There are some additional parameters such as plotting it for the full sensor or a single pixel.

As far as I can tell, yes. And also your coordinate transformation is perfectly fine.

I hope this helps - but your e-field definitely seems in order.

Best,
Simon