Inquiry on Using Allpix for Simulation of a 4x4 Pixel Detector with Weighting Potential and Electric Field

[fastjoy]

Hello,

I am a beginner in using Allpix, and I would like to simulate a 4x4 pixel detector that I have designed. I have already performed simulations in TCAD and obtained the weighting potential distribution for one of the electrodes (meaning I have selected the weighting potential distribution for a single electrode across the entire 4x4 detector) as well as the overall electric field distribution.

If I want to proceed with the simulation, is it correct to directly import the overall weighting potential and electric field? If so, could you please provide some examples related to the distribution of weighting potentials?

Thank you very much!

[pschutze]

Dear @fastjoy ,

if you already performed the TCAD simulations, you made about 80% of the way to your simulation, I would say. Note that we currently support the TCAD DF-ISE data format as well as output generated from Silvaco TCAD.

In short, you first need to export the fields from TCAD and convert them to an Allpix Squared compatible format. We provide a tool for converting these files which is documented here.

When having converted the fields, you can include the weighting potential reader (documentation here) and the electric field reader (similar syntax, documentation here) into your Allpix Squared configuration.

Please also have a look at this chapter of our documentation, which gives hints on which field map geometries are applicable and how to load them. In any case, you can enable control graphs (output_plots = true) for the corresponding field reader modules and validate the orientation of the loaded fields.

Please let us know in case you need further directions.

Cheers
Paul

[fastjoy]

sorry,this the is weight potential that i get from the simulation of the TCAD,

if allpix allow me to use it as PIXEL_QUADRANT_II in the simulation ，
For Allpix, is it necessary to simulate a single pixel in order to run a TCAD project? In my current simulation of 59.5 keV gamma, no peak appears, regardless of whether I import the potential or only the electric field.
it is my conf and the electric field

tcad_field_simulation.conf (1.9 KB)
weightpotential_convert.conf (444 Bytes)

[fastjoy]

Dear @pschutze ,
Aside from the previous questions, may I ask if Allpix is capable of extracting the signal readout for an individual pixel in order to determine its energy resolution?

[pschutze]

Dear @fastjoy ,

for the electric field, the imported field looks ok, but the configuration does not, if I’m not mistaken. I believe that the correct option would be to use the following parameters:

field_mapping = SENSOR
field_scale = 4.0, 4.0

The latter indicates, that the field spans four pixels.

As for the weighting potential, I would advise using one of the four central pixels as reference pixel for the TCAD simulation. However, to be fair, I’m not sure how well we can integrate asymmetric weighting potentials. This is maybe a question for @simonspa and @hwennlof .

I would advise you to use the Mesh Plotter tool and to enable the output_plots in the WeightingPotentialReader and the ElectricFieldReader module. These will help you to understand, which parameters you have to use to align your imported fields with the sensor.

To your other questions:

For Allpix, is it necessary to simulate a single pixel in order to run a TCAD project? In my current simulation of 59.5 keV gamma, no peak appears, regardless of whether I import the potential or only the electric field.

No, you can simulate a full grid of pixels. The reason for the peak not to appear must lie somewhere else and usually it’s a matter of the right configurations for loading the fields.

may I ask if Allpix is capable of extracting the signal readout for an individual pixel in order to determine its energy resolution?

Yes, and you already did the right thing by using the DetectorHistogrammer module, which should contain a graph named pixel_charge, and the ROOTObjectWriter, which is the alternative option for exporting data in the ROOT TTree data format. Examples on how to analyse such data can be found here in C and python, depending on your preference.

Cheers
Paul

[fastjoy]

Thank you very much for your response and help.

[fastjoy]

Dear @simonspa,
Regarding how to correctly load the weighting potential distribution for a 4x4 pixel detector, particularly for the corner pixels, into your simulation setup, please refer to the images and settings mentioned earlier. I would greatly appreciate your guidance on this matter.

[simonspa]

Hi @fastjoy

sorry for the late answer - currently we cannot flip/rotate weighting potentials. I might be looking into how to best do this. but for now I would recommend using a symmetric weighting potential centered around the pixel you put on unit potential.

Best,
Simon

[simonspa]

Sorry, just to add: the reason this is currently not implemented is that it is very hard to get TCAD results properly respect boundary conditions in these cases - which is not trivial since you cannot use the standard periodic boundary condition set. Therefore we restricted the type of maps you can use to avoid unphysical results from this.

[fastjoy]

Dear @simonspa,

Sorry, I didn’t quite catch your point clearly. Below are my configuration files and the related weighting potential distribution. Could you please advise me on how to set them up correctly?

CZT.conf (64 Bytes)
tcad_field_simulation.conf (2.0 KB)
tcd2.conf (579 Bytes)

[simonspa]

Hi @fastjoy

you need exactly one weighting potential and one electric field. The weighting potential should be a symmetric field centered around your electrode at unit potential.

/Simon

[fastjoy]

I’m very sorry for the delayed response. The reason for this delay is that I haven’t fully understood the relevant configuration. Based on my current understanding of the module, it seems that when reading the weighted potential field, it’s challenging to apply such a 4x4 structure properly across the entire system. From the examples I’ve seen, the provided weighted potentials are for individual pixel cells rather than an overall detector structure like mine.

Additionally, when I tried to apply this myself, the final results saved in the root file of the module didn’t quite meet my expectations. So, I haven’t been able to fully grasp your explanation.

I now plan to spend some time learning and attempt to simulate a single pixel cell in TCAD. Could you please clarify if Allpix has limitations when applying the simulation of a single pixel cell to an entire detector system? Or does Allpix expect me to provide the simulation results for a single pixel cell?

I would greatly appreciate your insights and apologize once again for the inconvenience.

[simonspa]

Hi @fastjoy

when reading the weighted potential field, it’s challenging to apply such a 4x4

No, it’s not challenging, but your simulations posted above will not produce physical results.

the provided weighted potentials are for individual pixel cells rather than an overall detector structure like mine.

Our fields almost never are for the entire detector, because they show a certain periodicity which we can easily simulate in Allpix Squared but that take very long in TCAD to simulate. THat is why for the electric field often only a single pixel is simulated. For the weighting potential you need a summetric potential around one pixel. How large this is depends on how far your weighting potential reaches, in most cases, one additional pixel in every direction should suffice, i.e. a 3x3 pixel grid with the central pixel set to unit potential.

/Simon