Detector Configuration and Mesh Sizes in Allpix-Squared

Hi all,

I’ve recently been experimenting with different detector configurations and TCAD meshes in Allpix-Squared, and I have some questions about how the sizes of the simulated detectors are handled.

So, a simple detector model would probably contain some lines like this:

type = “monolithic”

number_of_pixels = 1 1
pixel_size = 50um 100um
sensor_thickness = 300um

My understanding of this is that the detector here will be a single pixel, but the pixel will be 50 microns wide in the x-direction, 100 microns wide in the y-direction and 300 microns deep in the z-direction. Finally, the “sensor_excess” and “sensor_excess_direction” parameters can be used to change the size further, but I haven’t been using it so it’s a bit beyond the scope of this post.

Now, when using a TCAD mesh, it has it’s own dimensions, right? I’ve noticed that sometimes my configuration complains when I give it a TCAD mesh that (apparently) is a different size or shape to the detector configuration, but that only results in (WARNING) level messages which don’t stop allpix from finishing the simulation. However, when I look the histograms showing something like the electric field magnitude after this, they have incorrect values for sizes of some axes. These aren’t even the values that I input into the model file but allpix complained were incorrect, they’re often new numbers entirely.

My first question is basically how can I use allpix to find the size and shape of the TCAD meshes I have? My ultimate goal is to study how parameters like charge collection efficiency behave when charges are deposited in different points in a strip and to do that I need to know that they will be placed where I want them to be, but the behaviour described above has given me some pause.

My second question is related to the above. Below I’ve attached a screenshot of a histogram showing an electric field mesh I’m using

This is a cross-section of several ITk strips at the edge of a sensor. I have this as a single TCAD mesh, but for my purposes I only really need a small subset of this mesh (the entire extent in y, but only the interval [-0.45,-0.35] in z).

My second question is: does allpix have any methods to consider just a subset of a mesh like this? As I’ve described above, changing the detector model parameters doesn’t do that, and moving the detector around doesn’t work either.


Hi @damir

This is correct. :slightly_smiling_face:

You translated the TCAD mesh using the Mesh Converter I guess? When converting, the different dimensions should be printed on screen. In Allpix Squared itself, the definition of the local coordinate system applies as described in the user manual. The DepositionPointCharge module can be used to drop e/h pairs at specific positions in the sensor.

Yes, this is possible. However, before answering this I would like to point out that your field is wrongly oriented. Allpix Squared requires the electrodes to be positioned at +z, but your sensor implants are at -y. This is quite common because some TCAD tools use y as the thickness orientation. During conversion you can adjust and flip your field using the xyz parameter. Please fix this.

Then to answer your question: yes, this is possible. You can scale your field along z by setting

depletion_depth = 10um

for example. However, the field will always start at the surface, and everything below the applied field (deeper than the dpecified depth) will be considered undepleted.

Best regards,