1. Analysis objectives and run measurement run breakdown.
Breakdown of measurements performed.xlsx
2. Change in CompEnOn window with setting (0-7)
The transfer functions for a single pixel at different CompEnOn settings are seen above. CompEn appears to be entirely off at 0, and partially outside of the response time for the CSA/filter at 7. A value of 1 appears to be reasonable. Step size is still to be quantified. 1 tick is 10 ns.
3. Change in CompEn window onset between pixels and different timing rates:
The transfer window for a few different pixels at different run and daq triggers can be seen above (CompEn = 1). The end of the CompEnOn window appears to have significant variation between pixels.
4. Energy dependence of the transfer function (rise time variation for the CSA+filter system)
The transfer window for a few different pixels in FM when exposed to a low and high photon signal. A variation in the start of the flat region is observed (energy dependence) that needs to be considered when timing in the detector. Still needs to be quantified.
5. Dual linear responses when not running DAQ trigger=run trigger
This issues is being investigated as a firmware bug.
The behaviour is consistent with the observation of these dark images where on the first one the dark is aligned with the incoming data and the second one there is a shift on the incoming data. The shift seems to be related to an input FIFO not being cleared while operating the triggers at different rates.
Linearities were observed for a few pixels at different run and daq trigger rate combos (Run/DAQ). If run trigger is not same as daq trigger we see a double linearity. The cause for this has been traced to a firmware problem (pixel assignment issue). The collected data can be used if we use every other frame.
6. Initial evaluation of switching ranges and relative gain ratios and gains:
6.1 switching points and dynamic range
Above is a comparison of two pixel responses to an intensity scan. we see that we are using only about 6-8000ADU of the dynamic range of the detector. There is a significant difference in the switching point (12,000 ADU vs. 14,000ADU). A stial map of the maximum ADU observed in AML-M (i.e. switching point): can be seen below:
6.2 Relative gain ratios between medium and low in AML, and relative gain maps for AML-M and AML-L
As seen in the plot above, the relative gain ratios between medium and low gains in the switching mode for all bonded pixels gives a median gain ratio between medium and low of 0.037 (~1/27), with a variation of 9%. The fitting could be contributing to the large spread and needs to be evaluated.
The spatial distribution of the relative gain ratios (low gain slope/high gain slope) can be seen above and shows distinct patterns.
Above is the relative gain map for the low gain in AML. The two plots are the same, just with the right one more 'zoomed in' in the gain range. The value here is the slope of the intensity scan, i.e. ePixHR ADU/Wave-8 ADU. Pattern observable when we zoom in.
Above is the relative gainmap for the medium gain of AML (ePixHR ADU/Wave-8 ADU). The spatial pattern in the relative gain is observable here. Switching point position per pixel and fitting could have contributions and needs to be checked. Should be noted that the map directly follows the switching point map seen in section 6.1.
Performing bi-linear slope evaluation on FM still shows the pattern, as seen above, giving more credence to this being real gain variation.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
