Implement output amplifier crosstalk model
Detectors (or detector systems) often have cross-talk between output amplifiers. This causes real signal that is incident on one detector output to appear on other outputs. During the voltage-voltage conversion stage would be a good place to implement this.
There are two types of crosstalk we are concerned about.
The first is DC crosstalk, where a fraction of the true signal in the original output is seen the other outputs. This can be implemented as a basic matrix with fractional values--for example, let's imagine a 2 output detector. 1% of the signal incident on output 1 is seen in output 2, and 1% of the signal incident on output 2 is seen in output 1 (it is symmetric). Then the coupling matrix, a, would be [[1,0.01],[0.01, 1]] and the extra signal is multiplied by this matrix and then is then added to the original signal. (NOTE: the matrix does not always have to be symmetric!)
The second type is AC crosstalk, where the derivative of the signal in the original output is seen in the other outputs. Like the previous example, you would have a coupling matrix that describes the amplitude of the various couplings between the outputs. However, in this case the crosstalk is the derivative of the signal seen in the pixels in the order the pixels are read. So the extra signal you add in each output would be 𝑋𝑘,𝑗,𝑖 = 𝑎𝑘,𝑗 (𝑆𝑘,𝑖 − 𝑆𝑘, 𝑖−1) where k and j are indices over the amplifiers and i is the pixel index. This is a bit complicated to describe with text only, but this paper provides a full explanation with plots: https://arxiv.org/abs/1808.00790
Tasks:
-
Ensure that the order that the pixels are read out is user-accessible and definable per detector output. -
Implement DC crosstalk model. User should be able to define a crosstalk matrix. -
Implement AC crosstalk model. User should be able to define a crosstalk matrix. -
Make the DC and AC models variable - the magnitude of the coupling depends on many things (e.g. read speed) so the user should be able to calibrate it.